credit_Fraud_AutoEncoder.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "zxC6plHHJWr9",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "95b9806c-e863-4c15-b953-8fd1b34470bf"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/gdrive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from collections import Counter\n",
        "from imblearn.over_sampling import SMOTE\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from keras import losses\n",
        "from scipy import stats\n",
        "import tensorflow as tf\n",
        "from keras.models import Model,load_model,Sequential\n",
        "from keras.layers import Input, Dense\n",
        "from keras.callbacks import ModelCheckpoint,TensorBoard\n",
        "from keras import regularizers, optimizers\n",
        "from sklearn.metrics import confusion_matrix,auc,roc_curve,roc_auc_score\n",
        "from statsmodels.graphics.mosaicplot import mosaic\n",
        "from itertools import product\n",
        "from sklearn.compose import ColumnTransformer\n",
        "from sklearn.preprocessing import OneHotEncoder\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.pipeline import Pipeline\n",
        "import datetime as dt\n",
        "from sklearn import preprocessing\n",
        "import seaborn as sns\n",
        "from tensorflow.keras.optimizers import Adam\n",
        "from tensorflow.keras.callbacks import ModelCheckpoint, Callback\n",
        "\n",
        "sns.set(style=\"whitegrid\")\n",
        "np.random.seed(203)\n",
        "\n",
        "train_data = pd.read_csv('/content/gdrive/MyDrive/fraudTrain.csv')\n",
        "test_data = pd.read_csv('/content/gdrive/MyDrive/fraudTest.csv')\n",
        "train_data.dropna(inplace=True)\n",
        "test_data.dropna(inplace=True)\n",
        "train=train_data\n",
        "test=test_data"
      ],
      "metadata": {
        "id": "zGXBL1eTJccN"
      },
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train['month'] = pd.to_datetime(train['trans_date_trans_time']).dt.month\n",
        "train['day'] = pd.to_datetime(train['trans_date_trans_time']).dt.dayofweek\n",
        "train['hour'] = pd.to_datetime(train['trans_date_trans_time']).dt.hour\n",
        "train['age'] = dt.date.today().year - pd.to_datetime(train['dob']).dt.year\n",
        "train_new = train[['category','state' ,'amt','zip', 'lat', 'long', 'city_pop','merch_lat','merch_long','age', 'hour', 'day', 'month', 'is_fraud']]\n",
        "# train"
      ],
      "metadata": {
        "id": "SP9gLoqBJ1Se"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "test['month'] = pd.to_datetime(test['trans_date_trans_time']).dt.month\n",
        "test['day'] = pd.to_datetime(test['trans_date_trans_time']).dt.dayofweek\n",
        "test['hour'] = pd.to_datetime(test['trans_date_trans_time']).dt.hour\n",
        "test['age'] = dt.date.today().year - pd.to_datetime(test['dob']).dt.year\n",
        "test_new = test[['category','state','amt','zip', 'lat', 'long','city_pop','merch_lat','merch_long','age', 'hour', 'day', 'month', 'is_fraud']]"
      ],
      "metadata": {
        "id": "U4pcX2JhJ4Up"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_new.head()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 226
        },
        "id": "QUOevJZqyfr9",
        "outputId": "bb565658-b23c-46f9-92be-18fa35bd169c"
      },
      "execution_count": 64,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "        category state     amt    zip      lat      long  city_pop  merch_lat  \\\n",
              "0       misc_net    NC    4.97  28654  36.0788  -81.1781      3495  36.011293   \n",
              "1    grocery_pos    WA  107.23  99160  48.8878 -118.2105       149  49.159047   \n",
              "2  entertainment    ID  220.11  83252  42.1808 -112.2620      4154  43.150704   \n",
              "3  gas_transport    MT   45.00  59632  46.2306 -112.1138      1939  47.034331   \n",
              "4       misc_pos    VA   41.96  24433  38.4207  -79.4629        99  38.674999   \n",
              "\n",
              "   merch_long  age  hour  day  month  is_fraud  \n",
              "0  -82.048315   35     0    1      1       0.0  \n",
              "1 -118.186462   45     0    1      1       0.0  \n",
              "2 -112.154481   61     0    1      1       0.0  \n",
              "3 -112.561071   56     0    1      1       0.0  \n",
              "4  -78.632459   37     0    1      1       0.0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-68db2280-c5f5-464f-b0af-c5ca65c51e70\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>category</th>\n",
              "      <th>state</th>\n",
              "      <th>amt</th>\n",
              "      <th>zip</th>\n",
              "      <th>lat</th>\n",
              "      <th>long</th>\n",
              "      <th>city_pop</th>\n",
              "      <th>merch_lat</th>\n",
              "      <th>merch_long</th>\n",
              "      <th>age</th>\n",
              "      <th>hour</th>\n",
              "      <th>day</th>\n",
              "      <th>month</th>\n",
              "      <th>is_fraud</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>misc_net</td>\n",
              "      <td>NC</td>\n",
              "      <td>4.97</td>\n",
              "      <td>28654</td>\n",
              "      <td>36.0788</td>\n",
              "      <td>-81.1781</td>\n",
              "      <td>3495</td>\n",
              "      <td>36.011293</td>\n",
              "      <td>-82.048315</td>\n",
              "      <td>35</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>grocery_pos</td>\n",
              "      <td>WA</td>\n",
              "      <td>107.23</td>\n",
              "      <td>99160</td>\n",
              "      <td>48.8878</td>\n",
              "      <td>-118.2105</td>\n",
              "      <td>149</td>\n",
              "      <td>49.159047</td>\n",
              "      <td>-118.186462</td>\n",
              "      <td>45</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>entertainment</td>\n",
              "      <td>ID</td>\n",
              "      <td>220.11</td>\n",
              "      <td>83252</td>\n",
              "      <td>42.1808</td>\n",
              "      <td>-112.2620</td>\n",
              "      <td>4154</td>\n",
              "      <td>43.150704</td>\n",
              "      <td>-112.154481</td>\n",
              "      <td>61</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>gas_transport</td>\n",
              "      <td>MT</td>\n",
              "      <td>45.00</td>\n",
              "      <td>59632</td>\n",
              "      <td>46.2306</td>\n",
              "      <td>-112.1138</td>\n",
              "      <td>1939</td>\n",
              "      <td>47.034331</td>\n",
              "      <td>-112.561071</td>\n",
              "      <td>56</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>misc_pos</td>\n",
              "      <td>VA</td>\n",
              "      <td>41.96</td>\n",
              "      <td>24433</td>\n",
              "      <td>38.4207</td>\n",
              "      <td>-79.4629</td>\n",
              "      <td>99</td>\n",
              "      <td>38.674999</td>\n",
              "      <td>-78.632459</td>\n",
              "      <td>37</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>1</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-68db2280-c5f5-464f-b0af-c5ca65c51e70')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-68db2280-c5f5-464f-b0af-c5ca65c51e70 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-68db2280-c5f5-464f-b0af-c5ca65c51e70');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-181db59f-651d-41b5-9483-00b2d36586ff\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-181db59f-651d-41b5-9483-00b2d36586ff')\"\n",
              "            title=\"Suggest charts.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-181db59f-651d-41b5-9483-00b2d36586ff button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# all rows and all columns (except the last column)\n",
        "creditCardData = train_new.iloc[:, :-1].values\n",
        "# all rows and the last column => NOTE: storing all the result value\n",
        "creditCardLabels = train_new.iloc[:, -1].values"
      ],
      "metadata": {
        "id": "pd-AigKVKBNS"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "creditCardData.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vVbPjIjBu9hy",
        "outputId": "b1d0b0a1-fe5b-433f-b3e6-33fcb8f072fb"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(15592, 13)"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "categorical_columns = [0,1]\n",
        "numeric_columns = [2,3, 4, 5, 6, 7,8,9 , 10 ,11, 12]\n",
        "\n",
        "# Use one-hot encoding for the 'category' column\n",
        "preprocessor = ColumnTransformer([\n",
        "    ('num', StandardScaler(), numeric_columns),\n",
        "    ('cat', OneHotEncoder(sparse_output=False, drop='first'), categorical_columns)\n",
        "])\n",
        "# Apply preprocessing\n",
        "creditCardFeatures = preprocessor.fit_transform(creditCardData)"
      ],
      "metadata": {
        "id": "oducfEhdKmb4"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "creditCardFeatures.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "RMZF4a1-u7My",
        "outputId": "cfa6ccae-4f39-4d1a-f3cf-3e513784e75e"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(15592, 73)"
            ]
          },
          "metadata": {},
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#using SMOTE method to solve the imbalance problem\n",
        "sm = SMOTE(random_state=42)\n",
        "creditCardFeatures_sampled,creditCardLabels_sampled=sm.fit_resample(creditCardFeatures,creditCardLabels)"
      ],
      "metadata": {
        "id": "AHHKryxcQWPS"
      },
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "counts=pd.value_counts(creditCardLabels_sampled)\n",
        "print(counts)"
      ],
      "metadata": {
        "id": "hPhldaC4QvUH",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2e12848e-16de-4fb3-89f7-cd4ab8ad87af"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.0    15492\n",
            "1.0    15492\n",
            "dtype: int64\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "###Building Model"
      ],
      "metadata": {
        "id": "zBTh69CQQ92H"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "# <0.8 is a comparison that checks if the generated random numbers are less than 0.8.\n",
        "#This comparison results in a boolean array where each element is True if the corresponding\n",
        "#random number is less than 0.8 and False otherwise.\n",
        "\n",
        "#So, we are randomly assigning 80% of your data points to the training set\n",
        "# and the remaining 20% to the test set based on these randomly generated numbers.\n",
        "\n",
        "test_train_msk = np.random.rand(len(creditCardLabels_sampled)) < 0.8\n",
        "\n",
        "x_train = creditCardFeatures_sampled[test_train_msk]\n",
        "y_train = creditCardLabels_sampled[test_train_msk]\n",
        "\n",
        "x_test = creditCardFeatures_sampled[~test_train_msk]\n",
        "y_test = creditCardLabels_sampled[~test_train_msk]"
      ],
      "metadata": {
        "id": "lCEwbtnzQWRi"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_split_msk = np.random.rand(len(x_train)) < 0.5\n",
        "\n",
        "x_train_unsupervised = x_train[train_split_msk]\n",
        "y_train_unsupervised = y_train[train_split_msk]\n",
        "\n",
        "x_train_supervised = x_train[~train_split_msk]\n",
        "y_train_supervised = y_train[~train_split_msk]"
      ],
      "metadata": {
        "id": "zbMfRxQTRoX-"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "counts1= pd.value_counts(y_train_supervised)\n",
        "print(counts1)"
      ],
      "metadata": {
        "id": "CPfGV4rkSCkk",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "b9b4afdd-4dd3-42d5-fa21-d0bbae2ec70b"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1.0    6202\n",
            "0.0    6147\n",
            "dtype: int64\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_test.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "5jzJO-jjumCI",
        "outputId": "5718f2b4-cccd-4931-b9f9-8859e32580a9"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(6177, 73)"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_train_unsupervised.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oL7vpBouqd-N",
        "outputId": "7ebff543-5e1e-437f-c0a0-9a9270431564"
      },
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(12458, 73)"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_train_supervised.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qveUZX_Ws8zU",
        "outputId": "3e2263e0-b70d-4525-c055-2d659d91f550"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(12349, 73)"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#vverbose = A value of 1 means that you will receive progress messages about the training process.\n",
        "# save_best_only= When set to True, this ensures that the model weights are saved only when the monitored metric (in this case, validation accuracy) improves.\n",
        "# mode =  it's set to 'max' because we want to save the model with the highest validation accuracy.\n",
        "checkpoint = ModelCheckpoint('frauddetectionmodel.h5',monitor='val_acc',verbose=1,save_best_only=True,mode='max')\n",
        "\n",
        "#initialally, we have 73 features\n",
        "input_rec=Input(shape= (73,))\n",
        "#Hidden layers\n",
        "\n",
        "encoder = Dense(37,activation ='tanh')(input_rec)  #dense layer 1 in encoder\n",
        "encoder = Dense(18,activation ='relu')(encoder)    #dense layer 2 in encoder\n",
        "decoder = Dense(37,activation ='tanh')(encoder)    #dense layer 1 in decoder\n",
        "decoder = Dense(73,activation ='relu')(decoder)    #dense layer 2 in decoder\n",
        "\n",
        "encoder = Model(input_rec,encoder)        #Encoder model. we are not using this.\n",
        "#Build Model\n",
        "autoencoder = Model(input_rec, decoder)   #Full Model\n",
        "\n",
        "#using RMSPROP as optimizer\n",
        "rmsprop = tf.keras.optimizers.legacy.RMSprop(learning_rate=0.001,rho=0.9,epsilon=None,decay=0.0)\n",
        "# adam = Adam(learning_rate=0.001)\n",
        "\n",
        "autoencoder.compile(loss=losses.mean_squared_error,\n",
        "                    optimizer=rmsprop,\n",
        "                    metrics=['mse','acc'])\n",
        "\n",
        "#Training\n",
        "history = autoencoder.fit(\n",
        "    x_train_unsupervised,x_train_unsupervised,\n",
        "    batch_size=30,\n",
        "    shuffle=True,\n",
        "    callbacks=[checkpoint],\n",
        "    epochs=200,\n",
        "    validation_data=(x_test,x_test))"
      ],
      "metadata": {
        "id": "fpmLNeWUQWT_",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "97475c5b-74f5-4717-c7e2-74704eee7827"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1720 - mse: 0.1720 - acc: 0.6046\n",
            "Epoch 1: val_acc improved from -inf to 0.75409, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 10s 16ms/step - loss: 0.1717 - mse: 0.1717 - acc: 0.6074 - val_loss: 0.1307 - val_mse: 0.1307 - val_acc: 0.7541\n",
            "Epoch 2/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1284 - mse: 0.1284 - acc: 0.7833\n",
            "Epoch 2: val_acc improved from 0.75409 to 0.79246, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 3s 7ms/step - loss: 0.1282 - mse: 0.1282 - acc: 0.7843 - val_loss: 0.1247 - val_mse: 0.1247 - val_acc: 0.7925\n",
            "Epoch 3/200\n",
            "404/416 [============================>.] - ETA: 0s - loss: 0.1239 - mse: 0.1239 - acc: 0.8225\n",
            "Epoch 3: val_acc improved from 0.79246 to 0.83779, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1239 - mse: 0.1239 - acc: 0.8233 - val_loss: 0.1219 - val_mse: 0.1219 - val_acc: 0.8378\n",
            "Epoch 4/200\n",
            "397/416 [===========================>..] - ETA: 0s - loss: 0.1219 - mse: 0.1219 - acc: 0.8510\n",
            "Epoch 4: val_acc improved from 0.83779 to 0.85284, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1219 - mse: 0.1219 - acc: 0.8524 - val_loss: 0.1202 - val_mse: 0.1202 - val_acc: 0.8528\n",
            "Epoch 5/200\n",
            "409/416 [============================>.] - ETA: 0s - loss: 0.1207 - mse: 0.1207 - acc: 0.8620\n",
            "Epoch 5: val_acc improved from 0.85284 to 0.86434, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1207 - mse: 0.1207 - acc: 0.8619 - val_loss: 0.1193 - val_mse: 0.1193 - val_acc: 0.8643\n",
            "Epoch 6/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1198 - mse: 0.1198 - acc: 0.8663\n",
            "Epoch 6: val_acc did not improve from 0.86434\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1199 - mse: 0.1199 - acc: 0.8659 - val_loss: 0.1193 - val_mse: 0.1193 - val_acc: 0.8494\n",
            "Epoch 7/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1195 - mse: 0.1195 - acc: 0.8666\n",
            "Epoch 7: val_acc did not improve from 0.86434\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1195 - mse: 0.1195 - acc: 0.8666 - val_loss: 0.1184 - val_mse: 0.1184 - val_acc: 0.8588\n",
            "Epoch 8/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1196 - mse: 0.1196 - acc: 0.8654\n",
            "Epoch 8: val_acc did not improve from 0.86434\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1192 - mse: 0.1192 - acc: 0.8654 - val_loss: 0.1184 - val_mse: 0.1184 - val_acc: 0.8537\n",
            "Epoch 9/200\n",
            "399/416 [===========================>..] - ETA: 0s - loss: 0.1186 - mse: 0.1186 - acc: 0.8642\n",
            "Epoch 9: val_acc did not improve from 0.86434\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1190 - mse: 0.1190 - acc: 0.8646 - val_loss: 0.1190 - val_mse: 0.1190 - val_acc: 0.8592\n",
            "Epoch 10/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1191 - mse: 0.1191 - acc: 0.8614\n",
            "Epoch 10: val_acc did not improve from 0.86434\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1188 - mse: 0.1188 - acc: 0.8611 - val_loss: 0.1179 - val_mse: 0.1179 - val_acc: 0.8642\n",
            "Epoch 11/200\n",
            "404/416 [============================>.] - ETA: 0s - loss: 0.1186 - mse: 0.1186 - acc: 0.8652\n",
            "Epoch 11: val_acc did not improve from 0.86434\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1188 - mse: 0.1188 - acc: 0.8657 - val_loss: 0.1177 - val_mse: 0.1177 - val_acc: 0.8614\n",
            "Epoch 12/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1189 - mse: 0.1189 - acc: 0.8640\n",
            "Epoch 12: val_acc did not improve from 0.86434\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1187 - mse: 0.1187 - acc: 0.8651 - val_loss: 0.1180 - val_mse: 0.1180 - val_acc: 0.8540\n",
            "Epoch 13/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1186 - mse: 0.1186 - acc: 0.8643\n",
            "Epoch 13: val_acc did not improve from 0.86434\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1186 - mse: 0.1186 - acc: 0.8643 - val_loss: 0.1194 - val_mse: 0.1194 - val_acc: 0.8358\n",
            "Epoch 14/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1188 - mse: 0.1188 - acc: 0.8642\n",
            "Epoch 14: val_acc did not improve from 0.86434\n",
            "416/416 [==============================] - 3s 6ms/step - loss: 0.1186 - mse: 0.1186 - acc: 0.8638 - val_loss: 0.1180 - val_mse: 0.1180 - val_acc: 0.8443\n",
            "Epoch 15/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1185 - mse: 0.1185 - acc: 0.8661\n",
            "Epoch 15: val_acc improved from 0.86434 to 0.87211, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1185 - mse: 0.1185 - acc: 0.8666 - val_loss: 0.1176 - val_mse: 0.1176 - val_acc: 0.8721\n",
            "Epoch 16/200\n",
            "409/416 [============================>.] - ETA: 0s - loss: 0.1183 - mse: 0.1183 - acc: 0.8706\n",
            "Epoch 16: val_acc did not improve from 0.87211\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1184 - mse: 0.1184 - acc: 0.8707 - val_loss: 0.1179 - val_mse: 0.1179 - val_acc: 0.8619\n",
            "Epoch 17/200\n",
            "401/416 [===========================>..] - ETA: 0s - loss: 0.1183 - mse: 0.1183 - acc: 0.8713\n",
            "Epoch 17: val_acc did not improve from 0.87211\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1184 - mse: 0.1184 - acc: 0.8703 - val_loss: 0.1181 - val_mse: 0.1181 - val_acc: 0.8611\n",
            "Epoch 18/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1183 - mse: 0.1183 - acc: 0.8702\n",
            "Epoch 18: val_acc did not improve from 0.87211\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1183 - mse: 0.1183 - acc: 0.8704 - val_loss: 0.1174 - val_mse: 0.1174 - val_acc: 0.8718\n",
            "Epoch 19/200\n",
            "413/416 [============================>.] - ETA: 0s - loss: 0.1183 - mse: 0.1183 - acc: 0.8669\n",
            "Epoch 19: val_acc did not improve from 0.87211\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1183 - mse: 0.1183 - acc: 0.8671 - val_loss: 0.1181 - val_mse: 0.1181 - val_acc: 0.8601\n",
            "Epoch 20/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.8731\n",
            "Epoch 20: val_acc did not improve from 0.87211\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1182 - mse: 0.1182 - acc: 0.8729 - val_loss: 0.1174 - val_mse: 0.1174 - val_acc: 0.8624\n",
            "Epoch 21/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1182 - mse: 0.1182 - acc: 0.8712\n",
            "Epoch 21: val_acc did not improve from 0.87211\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1182 - mse: 0.1182 - acc: 0.8712 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.8719\n",
            "Epoch 22/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.8722\n",
            "Epoch 22: val_acc did not improve from 0.87211\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1182 - mse: 0.1182 - acc: 0.8720 - val_loss: 0.1183 - val_mse: 0.1183 - val_acc: 0.8410\n",
            "Epoch 23/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.8715\n",
            "Epoch 23: val_acc improved from 0.87211 to 0.87923, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1182 - mse: 0.1182 - acc: 0.8720 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8792\n",
            "Epoch 24/200\n",
            "404/416 [============================>.] - ETA: 0s - loss: 0.1184 - mse: 0.1184 - acc: 0.8702\n",
            "Epoch 24: val_acc did not improve from 0.87923\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1182 - mse: 0.1182 - acc: 0.8712 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.8695\n",
            "Epoch 25/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1182 - mse: 0.1182 - acc: 0.8727\n",
            "Epoch 25: val_acc did not improve from 0.87923\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1181 - mse: 0.1181 - acc: 0.8733 - val_loss: 0.1181 - val_mse: 0.1181 - val_acc: 0.8724\n",
            "Epoch 26/200\n",
            "404/416 [============================>.] - ETA: 0s - loss: 0.1186 - mse: 0.1186 - acc: 0.8716\n",
            "Epoch 26: val_acc did not improve from 0.87923\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1181 - mse: 0.1181 - acc: 0.8722 - val_loss: 0.1182 - val_mse: 0.1182 - val_acc: 0.8601\n",
            "Epoch 27/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.8734\n",
            "Epoch 27: val_acc did not improve from 0.87923\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1181 - mse: 0.1181 - acc: 0.8738 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8726\n",
            "Epoch 28/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1184 - mse: 0.1184 - acc: 0.8743\n",
            "Epoch 28: val_acc did not improve from 0.87923\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1181 - mse: 0.1181 - acc: 0.8746 - val_loss: 0.1183 - val_mse: 0.1183 - val_acc: 0.8401\n",
            "Epoch 29/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.8748\n",
            "Epoch 29: val_acc did not improve from 0.87923\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1181 - mse: 0.1181 - acc: 0.8749 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8690\n",
            "Epoch 30/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8764\n",
            "Epoch 30: val_acc did not improve from 0.87923\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1180 - mse: 0.1180 - acc: 0.8760 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8750\n",
            "Epoch 31/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1182 - mse: 0.1182 - acc: 0.8744\n",
            "Epoch 31: val_acc improved from 0.87923 to 0.88166, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1181 - mse: 0.1181 - acc: 0.8744 - val_loss: 0.1201 - val_mse: 0.1201 - val_acc: 0.8817\n",
            "Epoch 32/200\n",
            "402/416 [===========================>..] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8733\n",
            "Epoch 32: val_acc did not improve from 0.88166\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1180 - mse: 0.1180 - acc: 0.8741 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.8808\n",
            "Epoch 33/200\n",
            "406/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8761\n",
            "Epoch 33: val_acc did not improve from 0.88166\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1180 - mse: 0.1180 - acc: 0.8755 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8728\n",
            "Epoch 34/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8774\n",
            "Epoch 34: val_acc did not improve from 0.88166\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1180 - mse: 0.1180 - acc: 0.8777 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8687\n",
            "Epoch 35/200\n",
            "406/416 [============================>.] - ETA: 0s - loss: 0.1182 - mse: 0.1182 - acc: 0.8761\n",
            "Epoch 35: val_acc did not improve from 0.88166\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1180 - mse: 0.1180 - acc: 0.8758 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8758\n",
            "Epoch 36/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.8767\n",
            "Epoch 36: val_acc did not improve from 0.88166\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1180 - mse: 0.1180 - acc: 0.8769 - val_loss: 0.1182 - val_mse: 0.1182 - val_acc: 0.8703\n",
            "Epoch 37/200\n",
            "397/416 [===========================>..] - ETA: 0s - loss: 0.1187 - mse: 0.1187 - acc: 0.8755\n",
            "Epoch 37: val_acc did not improve from 0.88166\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1180 - mse: 0.1180 - acc: 0.8758 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8626\n",
            "Epoch 38/200\n",
            "409/416 [============================>.] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.8791\n",
            "Epoch 38: val_acc did not improve from 0.88166\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1180 - mse: 0.1180 - acc: 0.8786 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8805\n",
            "Epoch 39/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.8765\n",
            "Epoch 39: val_acc improved from 0.88166 to 0.88522, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1180 - mse: 0.1180 - acc: 0.8762 - val_loss: 0.1177 - val_mse: 0.1177 - val_acc: 0.8852\n",
            "Epoch 40/200\n",
            "406/416 [============================>.] - ETA: 0s - loss: 0.1182 - mse: 0.1182 - acc: 0.8776\n",
            "Epoch 40: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8781 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8745\n",
            "Epoch 41/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8764\n",
            "Epoch 41: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8767 - val_loss: 0.1176 - val_mse: 0.1176 - val_acc: 0.8833\n",
            "Epoch 42/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.8756\n",
            "Epoch 42: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8757 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8817\n",
            "Epoch 43/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8785\n",
            "Epoch 43: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8784 - val_loss: 0.1175 - val_mse: 0.1175 - val_acc: 0.8826\n",
            "Epoch 44/200\n",
            "398/416 [===========================>..] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8805\n",
            "Epoch 44: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8800 - val_loss: 0.1175 - val_mse: 0.1175 - val_acc: 0.8605\n",
            "Epoch 45/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.8786\n",
            "Epoch 45: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8786 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8752\n",
            "Epoch 46/200\n",
            "395/416 [===========================>..] - ETA: 0s - loss: 0.1173 - mse: 0.1173 - acc: 0.8822\n",
            "Epoch 46: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8817 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8711\n",
            "Epoch 47/200\n",
            "406/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8789\n",
            "Epoch 47: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8787 - val_loss: 0.1180 - val_mse: 0.1180 - val_acc: 0.8750\n",
            "Epoch 48/200\n",
            "404/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8796\n",
            "Epoch 48: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8791 - val_loss: 0.1175 - val_mse: 0.1175 - val_acc: 0.8776\n",
            "Epoch 49/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.8801\n",
            "Epoch 49: val_acc did not improve from 0.88522\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8800 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8823\n",
            "Epoch 50/200\n",
            "401/416 [===========================>..] - ETA: 0s - loss: 0.1184 - mse: 0.1184 - acc: 0.8789\n",
            "Epoch 50: val_acc improved from 0.88522 to 0.89008, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8788 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8901\n",
            "Epoch 51/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8808\n",
            "Epoch 51: val_acc did not improve from 0.89008\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8808 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8804\n",
            "Epoch 52/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1183 - mse: 0.1183 - acc: 0.8816\n",
            "Epoch 52: val_acc did not improve from 0.89008\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8812 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.8810\n",
            "Epoch 53/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.8782\n",
            "Epoch 53: val_acc did not improve from 0.89008\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8789 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8836\n",
            "Epoch 54/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.8848\n",
            "Epoch 54: val_acc did not improve from 0.89008\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1179 - mse: 0.1179 - acc: 0.8850 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8773\n",
            "Epoch 55/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8834\n",
            "Epoch 55: val_acc did not improve from 0.89008\n",
            "416/416 [==============================] - 3s 7ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8834 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8787\n",
            "Epoch 56/200\n",
            "402/416 [===========================>..] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.8810\n",
            "Epoch 56: val_acc did not improve from 0.89008\n",
            "416/416 [==============================] - 3s 6ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8814 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8779\n",
            "Epoch 57/200\n",
            "404/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8833\n",
            "Epoch 57: val_acc improved from 0.89008 to 0.89526, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8835 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8953\n",
            "Epoch 58/200\n",
            "399/416 [===========================>..] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8835\n",
            "Epoch 58: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8823 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8883\n",
            "Epoch 59/200\n",
            "396/416 [===========================>..] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8868\n",
            "Epoch 59: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8872 - val_loss: 0.1175 - val_mse: 0.1175 - val_acc: 0.8880\n",
            "Epoch 60/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8858\n",
            "Epoch 60: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8853 - val_loss: 0.1187 - val_mse: 0.1187 - val_acc: 0.8919\n",
            "Epoch 61/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8837\n",
            "Epoch 61: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8841 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8862\n",
            "Epoch 62/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.8820\n",
            "Epoch 62: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8823 - val_loss: 0.1174 - val_mse: 0.1174 - val_acc: 0.8760\n",
            "Epoch 63/200\n",
            "396/416 [===========================>..] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8817\n",
            "Epoch 63: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8832 - val_loss: 0.1174 - val_mse: 0.1174 - val_acc: 0.8849\n",
            "Epoch 64/200\n",
            "413/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8856\n",
            "Epoch 64: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8858 - val_loss: 0.1177 - val_mse: 0.1177 - val_acc: 0.8875\n",
            "Epoch 65/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8877\n",
            "Epoch 65: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8874 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8912\n",
            "Epoch 66/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8864\n",
            "Epoch 66: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8864 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8810\n",
            "Epoch 67/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1183 - mse: 0.1183 - acc: 0.8847\n",
            "Epoch 67: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8847 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8736\n",
            "Epoch 68/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8862\n",
            "Epoch 68: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8862 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8885\n",
            "Epoch 69/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8888\n",
            "Epoch 69: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8886 - val_loss: 0.1176 - val_mse: 0.1176 - val_acc: 0.8690\n",
            "Epoch 70/200\n",
            "413/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8868\n",
            "Epoch 70: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8871 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8849\n",
            "Epoch 71/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1172 - mse: 0.1172 - acc: 0.8887\n",
            "Epoch 71: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8888 - val_loss: 0.1180 - val_mse: 0.1180 - val_acc: 0.8807\n",
            "Epoch 72/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.8873\n",
            "Epoch 72: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8865 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8899\n",
            "Epoch 73/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8858\n",
            "Epoch 73: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8859 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8740\n",
            "Epoch 74/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8875\n",
            "Epoch 74: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8877 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8825\n",
            "Epoch 75/200\n",
            "409/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8867\n",
            "Epoch 75: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8863 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8810\n",
            "Epoch 76/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1173 - mse: 0.1173 - acc: 0.8868\n",
            "Epoch 76: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8870 - val_loss: 0.1174 - val_mse: 0.1174 - val_acc: 0.8708\n",
            "Epoch 77/200\n",
            "399/416 [===========================>..] - ETA: 0s - loss: 0.1173 - mse: 0.1173 - acc: 0.8890\n",
            "Epoch 77: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8885 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8804\n",
            "Epoch 78/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8869\n",
            "Epoch 78: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8869 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8778\n",
            "Epoch 79/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8872\n",
            "Epoch 79: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8871 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8925\n",
            "Epoch 80/200\n",
            "409/416 [============================>.] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.8901\n",
            "Epoch 80: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8910 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8851\n",
            "Epoch 81/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.8892\n",
            "Epoch 81: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8891 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8781\n",
            "Epoch 82/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8900\n",
            "Epoch 82: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8900 - val_loss: 0.1174 - val_mse: 0.1174 - val_acc: 0.8865\n",
            "Epoch 83/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8891\n",
            "Epoch 83: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8895 - val_loss: 0.1177 - val_mse: 0.1177 - val_acc: 0.8839\n",
            "Epoch 84/200\n",
            "395/416 [===========================>..] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8906\n",
            "Epoch 84: val_acc did not improve from 0.89526\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8908 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8888\n",
            "Epoch 85/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.8872\n",
            "Epoch 85: val_acc improved from 0.89526 to 0.89655, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1178 - mse: 0.1178 - acc: 0.8879 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8966\n",
            "Epoch 86/200\n",
            "398/416 [===========================>..] - ETA: 0s - loss: 0.1187 - mse: 0.1187 - acc: 0.8892\n",
            "Epoch 86: val_acc did not improve from 0.89655\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8889 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8914\n",
            "Epoch 87/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8913\n",
            "Epoch 87: val_acc did not improve from 0.89655\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8910 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.8898\n",
            "Epoch 88/200\n",
            "398/416 [===========================>..] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8869\n",
            "Epoch 88: val_acc did not improve from 0.89655\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8877 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8886\n",
            "Epoch 89/200\n",
            "397/416 [===========================>..] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.8875\n",
            "Epoch 89: val_acc did not improve from 0.89655\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8884 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8898\n",
            "Epoch 90/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8901\n",
            "Epoch 90: val_acc did not improve from 0.89655\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8900 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.8595\n",
            "Epoch 91/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8922\n",
            "Epoch 91: val_acc did not improve from 0.89655\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8921 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8917\n",
            "Epoch 92/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8911\n",
            "Epoch 92: val_acc improved from 0.89655 to 0.89688, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8911 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8969\n",
            "Epoch 93/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8927\n",
            "Epoch 93: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8925 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8868\n",
            "Epoch 94/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8923\n",
            "Epoch 94: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8918 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8951\n",
            "Epoch 95/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8922\n",
            "Epoch 95: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8919 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8854\n",
            "Epoch 96/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.8941\n",
            "Epoch 96: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8940 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8771\n",
            "Epoch 97/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8904\n",
            "Epoch 97: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8904 - val_loss: 0.1182 - val_mse: 0.1182 - val_acc: 0.8561\n",
            "Epoch 98/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.8914\n",
            "Epoch 98: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8914 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8933\n",
            "Epoch 99/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8907\n",
            "Epoch 99: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8907 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8826\n",
            "Epoch 100/200\n",
            "406/416 [============================>.] - ETA: 0s - loss: 0.1173 - mse: 0.1173 - acc: 0.8938\n",
            "Epoch 100: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8938 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8867\n",
            "Epoch 101/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8945\n",
            "Epoch 101: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8945 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.8914\n",
            "Epoch 102/200\n",
            "399/416 [===========================>..] - ETA: 0s - loss: 0.1165 - mse: 0.1165 - acc: 0.8924\n",
            "Epoch 102: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8923 - val_loss: 0.1175 - val_mse: 0.1175 - val_acc: 0.8909\n",
            "Epoch 103/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8920\n",
            "Epoch 103: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8920 - val_loss: 0.1174 - val_mse: 0.1174 - val_acc: 0.8914\n",
            "Epoch 104/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1174 - mse: 0.1174 - acc: 0.8937\n",
            "Epoch 104: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8940 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8880\n",
            "Epoch 105/200\n",
            "402/416 [===========================>..] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.8944\n",
            "Epoch 105: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8947 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8891\n",
            "Epoch 106/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8958\n",
            "Epoch 106: val_acc did not improve from 0.89688\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8958 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8919\n",
            "Epoch 107/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8928\n",
            "Epoch 107: val_acc improved from 0.89688 to 0.90076, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8929 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.9008\n",
            "Epoch 108/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8965\n",
            "Epoch 108: val_acc did not improve from 0.90076\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8963 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8828\n",
            "Epoch 109/200\n",
            "409/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8954\n",
            "Epoch 109: val_acc did not improve from 0.90076\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8951 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8928\n",
            "Epoch 110/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8961\n",
            "Epoch 110: val_acc did not improve from 0.90076\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8959 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8885\n",
            "Epoch 111/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8986\n",
            "Epoch 111: val_acc improved from 0.90076 to 0.90303, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8987 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9030\n",
            "Epoch 112/200\n",
            "397/416 [===========================>..] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8966\n",
            "Epoch 112: val_acc improved from 0.90303 to 0.90772, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8961 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9077\n",
            "Epoch 113/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8950\n",
            "Epoch 113: val_acc did not improve from 0.90772\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8953 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8927\n",
            "Epoch 114/200\n",
            "399/416 [===========================>..] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8988\n",
            "Epoch 114: val_acc did not improve from 0.90772\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8991 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9046\n",
            "Epoch 115/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8979\n",
            "Epoch 115: val_acc did not improve from 0.90772\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8978 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8919\n",
            "Epoch 116/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8989\n",
            "Epoch 116: val_acc did not improve from 0.90772\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8981 - val_loss: 0.1175 - val_mse: 0.1175 - val_acc: 0.8902\n",
            "Epoch 117/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.8972\n",
            "Epoch 117: val_acc did not improve from 0.90772\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8971 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8959\n",
            "Epoch 118/200\n",
            "413/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8961\n",
            "Epoch 118: val_acc did not improve from 0.90772\n",
            "416/416 [==============================] - 3s 7ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8963 - val_loss: 0.1175 - val_mse: 0.1175 - val_acc: 0.8787\n",
            "Epoch 119/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8942\n",
            "Epoch 119: val_acc did not improve from 0.90772\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8942 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8925\n",
            "Epoch 120/200\n",
            "402/416 [===========================>..] - ETA: 0s - loss: 0.1186 - mse: 0.1186 - acc: 0.8983\n",
            "Epoch 120: val_acc did not improve from 0.90772\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.8980 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9069\n",
            "Epoch 121/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8969\n",
            "Epoch 121: val_acc did not improve from 0.90772\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8970 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9038\n",
            "Epoch 122/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.9005\n",
            "Epoch 122: val_acc improved from 0.90772 to 0.91031, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1177 - mse: 0.1177 - acc: 0.9009 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.9103\n",
            "Epoch 123/200\n",
            "409/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.8991\n",
            "Epoch 123: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8996 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9097\n",
            "Epoch 124/200\n",
            "402/416 [===========================>..] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8984\n",
            "Epoch 124: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8989 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8917\n",
            "Epoch 125/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8986\n",
            "Epoch 125: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8985 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.9040\n",
            "Epoch 126/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1174 - mse: 0.1174 - acc: 0.8969\n",
            "Epoch 126: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8966 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8975\n",
            "Epoch 127/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8993\n",
            "Epoch 127: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8994 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8956\n",
            "Epoch 128/200\n",
            "413/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.9000\n",
            "Epoch 128: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8998 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.9029\n",
            "Epoch 129/200\n",
            "404/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.8991\n",
            "Epoch 129: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8995 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.8996\n",
            "Epoch 130/200\n",
            "399/416 [===========================>..] - ETA: 0s - loss: 0.1172 - mse: 0.1172 - acc: 0.9002\n",
            "Epoch 130: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9000 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8993\n",
            "Epoch 131/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.9028\n",
            "Epoch 131: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9027 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8936\n",
            "Epoch 132/200\n",
            "396/416 [===========================>..] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.9009\n",
            "Epoch 132: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9005 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.9051\n",
            "Epoch 133/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.8983\n",
            "Epoch 133: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8979 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8998\n",
            "Epoch 134/200\n",
            "413/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9022\n",
            "Epoch 134: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9017 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.9004\n",
            "Epoch 135/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.9017\n",
            "Epoch 135: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9014 - val_loss: 0.1177 - val_mse: 0.1177 - val_acc: 0.8750\n",
            "Epoch 136/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1168 - mse: 0.1168 - acc: 0.9004\n",
            "Epoch 136: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9009 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.9102\n",
            "Epoch 137/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.9023\n",
            "Epoch 137: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9025 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9084\n",
            "Epoch 138/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.9007\n",
            "Epoch 138: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9004 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.9093\n",
            "Epoch 139/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.9019\n",
            "Epoch 139: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9017 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.9071\n",
            "Epoch 140/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.9016\n",
            "Epoch 140: val_acc did not improve from 0.91031\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9015 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8901\n",
            "Epoch 141/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.9016\n",
            "Epoch 141: val_acc improved from 0.91031 to 0.91889, saving model to frauddetectionmodel.h5\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9012 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.9189\n",
            "Epoch 142/200\n",
            "395/416 [===========================>..] - ETA: 0s - loss: 0.1183 - mse: 0.1183 - acc: 0.9041\n",
            "Epoch 142: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9028 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8969\n",
            "Epoch 143/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.9009\n",
            "Epoch 143: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9009 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9027\n",
            "Epoch 144/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.9004\n",
            "Epoch 144: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9005 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9056\n",
            "Epoch 145/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.8996\n",
            "Epoch 145: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8999 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.9017\n",
            "Epoch 146/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1181 - mse: 0.1181 - acc: 0.9026\n",
            "Epoch 146: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9017 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9085\n",
            "Epoch 147/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1171 - mse: 0.1171 - acc: 0.9009\n",
            "Epoch 147: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8997 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.9055\n",
            "Epoch 148/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1171 - mse: 0.1171 - acc: 0.9012\n",
            "Epoch 148: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9011 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8915\n",
            "Epoch 149/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1180 - mse: 0.1180 - acc: 0.9012\n",
            "Epoch 149: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9014 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9043\n",
            "Epoch 150/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.8984\n",
            "Epoch 150: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.8985 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9025\n",
            "Epoch 151/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9000\n",
            "Epoch 151: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9000 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.9043\n",
            "Epoch 152/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1171 - mse: 0.1171 - acc: 0.9036\n",
            "Epoch 152: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9035 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9001\n",
            "Epoch 153/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1171 - mse: 0.1171 - acc: 0.9028\n",
            "Epoch 153: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9034 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9048\n",
            "Epoch 154/200\n",
            "397/416 [===========================>..] - ETA: 0s - loss: 0.1184 - mse: 0.1184 - acc: 0.9018\n",
            "Epoch 154: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9022 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8962\n",
            "Epoch 155/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9040\n",
            "Epoch 155: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9042 - val_loss: 0.1176 - val_mse: 0.1176 - val_acc: 0.8967\n",
            "Epoch 156/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.9038\n",
            "Epoch 156: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9034 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8907\n",
            "Epoch 157/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9013\n",
            "Epoch 157: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9013 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8932\n",
            "Epoch 158/200\n",
            "413/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9023\n",
            "Epoch 158: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9022 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9076\n",
            "Epoch 159/200\n",
            "413/416 [============================>.] - ETA: 0s - loss: 0.1173 - mse: 0.1173 - acc: 0.9065\n",
            "Epoch 159: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9066 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8842\n",
            "Epoch 160/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.9020\n",
            "Epoch 160: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9022 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9058\n",
            "Epoch 161/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9049\n",
            "Epoch 161: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 3s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9050 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8962\n",
            "Epoch 162/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.9026\n",
            "Epoch 162: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9027 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9144\n",
            "Epoch 163/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1169 - mse: 0.1169 - acc: 0.9021\n",
            "Epoch 163: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9022 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8804\n",
            "Epoch 164/200\n",
            "414/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9015\n",
            "Epoch 164: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9013 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8998\n",
            "Epoch 165/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9053\n",
            "Epoch 165: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 3s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9059 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8983\n",
            "Epoch 166/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9016\n",
            "Epoch 166: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9017 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8998\n",
            "Epoch 167/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9038\n",
            "Epoch 167: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9042 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.9059\n",
            "Epoch 168/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.9041\n",
            "Epoch 168: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9039 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9136\n",
            "Epoch 169/200\n",
            "404/416 [============================>.] - ETA: 0s - loss: 0.1173 - mse: 0.1173 - acc: 0.9064\n",
            "Epoch 169: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9058 - val_loss: 0.1174 - val_mse: 0.1174 - val_acc: 0.8807\n",
            "Epoch 170/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9017\n",
            "Epoch 170: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9017 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.9045\n",
            "Epoch 171/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9036\n",
            "Epoch 171: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9037 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.8932\n",
            "Epoch 172/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1174 - mse: 0.1174 - acc: 0.9044\n",
            "Epoch 172: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9042 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8995\n",
            "Epoch 173/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.9060\n",
            "Epoch 173: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9060 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9040\n",
            "Epoch 174/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9052\n",
            "Epoch 174: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9050 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9025\n",
            "Epoch 175/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1172 - mse: 0.1172 - acc: 0.9061\n",
            "Epoch 175: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9057 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9129\n",
            "Epoch 176/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.9024\n",
            "Epoch 176: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9027 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8936\n",
            "Epoch 177/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9063\n",
            "Epoch 177: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9062 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.9071\n",
            "Epoch 178/200\n",
            "406/416 [============================>.] - ETA: 0s - loss: 0.1171 - mse: 0.1171 - acc: 0.9033\n",
            "Epoch 178: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9036 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.9050\n",
            "Epoch 179/200\n",
            "400/416 [===========================>..] - ETA: 0s - loss: 0.1172 - mse: 0.1172 - acc: 0.9026\n",
            "Epoch 179: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9019 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.9059\n",
            "Epoch 180/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1172 - mse: 0.1172 - acc: 0.9029\n",
            "Epoch 180: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9038 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.9046\n",
            "Epoch 181/200\n",
            "412/416 [============================>.] - ETA: 0s - loss: 0.1173 - mse: 0.1173 - acc: 0.9044\n",
            "Epoch 181: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9045 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.8755\n",
            "Epoch 182/200\n",
            "416/416 [==============================] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9045\n",
            "Epoch 182: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9045 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8941\n",
            "Epoch 183/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9042\n",
            "Epoch 183: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9046 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9016\n",
            "Epoch 184/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1173 - mse: 0.1173 - acc: 0.9031\n",
            "Epoch 184: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9026 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9158\n",
            "Epoch 185/200\n",
            "411/416 [============================>.] - ETA: 0s - loss: 0.1177 - mse: 0.1177 - acc: 0.9052\n",
            "Epoch 185: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9053 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8980\n",
            "Epoch 186/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1174 - mse: 0.1174 - acc: 0.9022\n",
            "Epoch 186: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 5ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9023 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9079\n",
            "Epoch 187/200\n",
            "415/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9048\n",
            "Epoch 187: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1176 - mse: 0.1176 - acc: 0.9048 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.8859\n",
            "Epoch 188/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9079\n",
            "Epoch 188: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9074 - val_loss: 0.1173 - val_mse: 0.1173 - val_acc: 0.8957\n",
            "Epoch 189/200\n",
            "409/416 [============================>.] - ETA: 0s - loss: 0.1172 - mse: 0.1172 - acc: 0.9065\n",
            "Epoch 189: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9068 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.8944\n",
            "Epoch 190/200\n",
            "406/416 [============================>.] - ETA: 0s - loss: 0.1179 - mse: 0.1179 - acc: 0.9072\n",
            "Epoch 190: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 4ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9072 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.9025\n",
            "Epoch 191/200\n",
            "395/416 [===========================>..] - ETA: 0s - loss: 0.1183 - mse: 0.1183 - acc: 0.9041\n",
            "Epoch 191: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9048 - val_loss: 0.1172 - val_mse: 0.1172 - val_acc: 0.9021\n",
            "Epoch 192/200\n",
            "403/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9055\n",
            "Epoch 192: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 1s 3ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9058 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.8956\n",
            "Epoch 193/200\n",
            "410/416 [============================>.] - ETA: 0s - loss: 0.1175 - mse: 0.1175 - acc: 0.9060\n",
            "Epoch 193: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9058 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9131\n",
            "Epoch 194/200\n",
            "413/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.9060\n",
            "Epoch 194: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 6ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9062 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.9032\n",
            "Epoch 195/200\n",
            "405/416 [============================>.] - ETA: 0s - loss: 0.1178 - mse: 0.1178 - acc: 0.9077\n",
            "Epoch 195: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9079 - val_loss: 0.1171 - val_mse: 0.1171 - val_acc: 0.8993\n",
            "Epoch 196/200\n",
            "396/416 [===========================>..] - ETA: 0s - loss: 0.1173 - mse: 0.1173 - acc: 0.9063\n",
            "Epoch 196: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9056 - val_loss: 0.1170 - val_mse: 0.1170 - val_acc: 0.8977\n",
            "Epoch 197/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1174 - mse: 0.1174 - acc: 0.9021\n",
            "Epoch 197: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9023 - val_loss: 0.1167 - val_mse: 0.1167 - val_acc: 0.8972\n",
            "Epoch 198/200\n",
            "408/416 [============================>.] - ETA: 0s - loss: 0.1171 - mse: 0.1171 - acc: 0.9032\n",
            "Epoch 198: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9037 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.9113\n",
            "Epoch 199/200\n",
            "407/416 [============================>.] - ETA: 0s - loss: 0.1176 - mse: 0.1176 - acc: 0.9020\n",
            "Epoch 199: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9022 - val_loss: 0.1169 - val_mse: 0.1169 - val_acc: 0.8930\n",
            "Epoch 200/200\n",
            "409/416 [============================>.] - ETA: 0s - loss: 0.1172 - mse: 0.1172 - acc: 0.9026\n",
            "Epoch 200: val_acc did not improve from 0.91889\n",
            "416/416 [==============================] - 2s 4ms/step - loss: 0.1175 - mse: 0.1175 - acc: 0.9024 - val_loss: 0.1168 - val_mse: 0.1168 - val_acc: 0.9074\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.utils import plot_model\n",
        "\n",
        "# Visualize the model and save it to a file\n",
        "plot_model(autoencoder, to_file='autoencoder.png', show_shapes=True, show_layer_names=True)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 533
        },
        "id": "mEUZndwX44b2",
        "outputId": "c638e522-ff04-4f53-cd63-1c40f92b6a85"
      },
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAASwAAAIECAIAAAAsCITkAAAABmJLR0QA/wD/AP+gvaeTAAAgAElEQVR4nOzdeVwT1944/jOBLCQhbLLJJgQV96VqBfFq9RYf9QVqwYqKdbn1xq2IoLW4UEVccrFCodpe11erVUDl4lKxbhetT9Xar1AUHxWxskhZBQImSgjz+2Nu55cbIIQkzEnw8/6LnJmcOWcmH2Y5M/MhSJJEAAB8WLgbAMDbDoIQAMwgCAHADIIQAMws1T/cunVrz549uJoCwFsiOjra39+f/vhfe8LS0tJTp04x3iTzdvv27du3b+NuRbcoKyuD34PRnTp1qrS0VL3Esu1MJ0+eZKo9PcHs2bNRD11pGRkZc+bM6ZFdw4ggCI0SOCcEADMIQgAwgyAEADMIQgAwgyAEALNuDMILFy7Y2NicO3eu+xaho/j4+IEDB4pEIi6X6+vr++mnnzY1NWFsj+msGaNYtmwZ8aeIiAj1SVeuXImNjT19+rSPjw81w4IFC9RnCAoKsra2trCwGDRo0L1795hs9sSJE4k2hEIhQkgqlfr5+VlZWQkEAj8/v82bN8tkMoTQ2bNnpVKpSqWiK8nKyqK/26tXL/1a0o1BaDrPZ1y7dm3VqlXPnz+vqanZsWNHcnIyNa6Ai+msGWOxt7fPzs5+/PjxoUOH6MLPP/88JSVlw4YNoaGhz549E4vFDg4Ox44d++GHH+h5Ll26dPLkyeDg4IKCgpEjR+Jo+38JDAxECP30009Lly4tKSmprKzctm2bVCoNCwtDCIWEhPB4vMmTJ9fX11Pzz5gxo6ys7MaNG9OmTdN/qaSa9PR0jRJTJpfL/f39dZlz+vTpLS0t9McPP/wQIVRSUmKUZoSFhYWFhRmlKmPRfc1op+PvQSKRuLm5aRTu3LmzX79+CoWCLhGLxd9//z2LxXJzc6uvr6fLs7OzZ8yYYXhru2rKlCkymUy9RCKRXL16lSTJWbNmqbec+pddXl5OfYyMjPT391cqlerfXb16tYODgy7LRQilp6erl5jxOeGhQ4eqqqp0mfP8+fMWFhb0R+qwQS6Xd1fLcNN9zXSTp0+fbt68eevWrTweT708ICAgKirqxYsXa9euxdU22sWLF62tremPpaWlDx48mDRpEkIoMzNTveVubm4IIfoUZsuWLXl5ecnJycZqSXcF4c2bNz09PQmC+OqrrxBC+/btEwgEfD7/zJkzU6dOFYlE7u7uJ06cQAilpKTweDwnJ6dly5a5urryeLyAgIA7d+4ghCIjIzkcjouLC1XnypUrBQIBQRA1NTVRUVExMTFFRUUEQfj6+napbS9evLCysvL29jZ2p3XC/Jq5ePGiSCTavn07Y31MSUkhSTIkJKTtpISEhH79+h08ePDKlSttp5IkuWfPngEDBnC5XDs7u5kzZz569AhpXUsIIZVKFRcX5+npaWVlNXToUGoH3lW7du1avXp1u5MKCwttbW29vLyoj3Z2dhMmTEhOTiaNdVqhvls07uEodYNcamoq9XHjxo0IoatXrzY0NFRVVY0fP14gEDQ3N5MkKZFIBALBw4cPX79+XVBQMHr0aGtra+pwcf78+c7OznSdiYmJCKHq6mqSJENDQ8VicVdb9erVK2tr68jISON0Uq/DUYbXzPnz562trePj47vaNb0PR318fAYOHKgxm1gs/v3330mS/Pnnn1ksVp8+fZqamsj/PhyNi4vjcDhHjx6tr6/Pz88fOXJkr169KioqtK+ltWvXcrncU6dO1dXVbdiwgcVi3b17t0s9LSsrGzhwoEqlUi9sbm4uKytLTU3lcrlHjx5VnxQbG4sQys3NpUvM6XA0ICBAJBI5OjqGh4e/evWqpKSEKre0tKT+/w0cOHDfvn2NjY1Hjhzpjgbs2LHD1dU1ISGhOyo3RPetmenTp8tkss2bN3dDq9vx6tWr33//XSwWdzSDv7//mjVrnj9//tlnn6mXKxSKPXv2fPDBBxERETY2NkOGDPnmm29qamr2799Pz9N2Lb1+/Xrfvn2zZs0KDQ21tbXdtGkTm83u6iratWvXJ598wmL9Vzh4eHi4u7tv2bLlH//4x5w5c9Qn9e3bFyF0//79Li2lI9jOCTkcDkJIqVS2nTRq1Cg+n08dhxhXZmZmRkbGjz/+qH4yYGqwrBkjqqqqIkmSz+drmSchIaF///579+69efMmXVhQUNDU1DRq1Ci6ZPTo0RwOhzoC10CvpcePH8vl8sGDB1PlVlZWLi4uXVpF5eXlZ8+eXbRokUZ5aWlpVVXV8ePHv/322xEjRqifZlO9q6ys1H0pWpjohRkul1tdXW3cOtPS0nbt2pWTk9OnTx/j1syk7lgzxvX69WuEEJfL1TIPj8c7cuQIQRBLlixRKBRUIXXdnxqpo9na2jY2Nmqp6tWrVwihTZs20eN1xcXFXbrqJpVKly5dqnENCSHEZrMdHR2DgoLS0tIKCgp27NhBT7KysqJ7ajhTDEKlUllfX+/u7m7EOlNTU48dO3bt2rXevXsbsVqGdceaMTrqB6o+ot0uf3//6OjowsLCbdu2USW2trYIIY2Q67S/jo6OCKGkpCT1s6xbt27p2NqKiorjx4+vWLFCyzy+vr4WFhYFBQV0SXNzM/qzp4YzxSDMyckhSXLs2LEIIUtLy3YPzHRHkuT69evv37+flZWl8V/W7Bh3zXQTJycngiAaGho6nXPbtm1+fn65ubnUx8GDBwuFwl9//ZWe4c6dO83Nze+8846WSjw8PHg8Xl5enn6tlUqlERER9vb2dEltbe28efPU5yksLFSpVB4eHnQJ1TtnZ2f9FqrBVIKwtbW1rq6upaUlPz8/KirK09OTOkb39fV9+fJlVlaWUqmsrq4uLi6mv2Jvb19eXv78+fPGxkYtP8eHDx/+4x//OHDgAJvNVr9Baffu3Qz0y3CGr5ns7Gwmhyj4fL6Pj09ZWVmnc1IHpfQQLo/Hi4mJyczMPHbsmEwmu3///vLly11dXSUSifZKFi9efOLEiX379slkMpVKVVZW9scffyCEwsPDnZ2dtdwNV1lZefjw4TVr1qgXCgSCS5cuXbt2TSaTKZXK3NzchQsXCgSC6Ohoeh6qd0OGDOm0jzpR34kbcYgiNTWVGsXi8/khISF79+6lzmX79u1bVFS0f/9+kUiEEPLy8nry5IlEImGz2W5ubpaWliKRaObMmUVFRVQ9tbW17733Ho/H8/b2/uSTT9atW4cQ8vX1LSkpuXfvnpeXl5WVVWBgIHUVu10dXcJKTEw0Sk+7OkTB/Jq5cOGCtbV1QkJCV7um9xBFZGQkm82Wy+XUx8zMTOpiaa9evVatWqXx9XXr1tFDFK2trYmJiX379mWz2XZ2drNmzXr8+DFJktrX0ps3b9avX+/p6Wlpaeno6BgaGlpQUECS5KxZsxBCcXFxHbU8Ojo6IiKibXlISIi3t7dQKORyuWKxODw8/P79++ozTJ8+3c3NrbW1lS4xZIjCJG5bk0gk9vb2zC/XKLr1tjW8a0bvICwsLLS0tNQYW2OeSqUaP378oUOHjFttTU0Nj8fbvXu3eqE5jRN2pNPz+LeWWawZhULx448/FhYWUlcsfH194+Pj4+PjMT6tolKpsrKyGhsbw8PDjVvzli1bhg8fHhkZiRAiSbK8vPzmzZtPnz7Vu0JTCUJDPHr0qO0zKTSjbwPQ1suXL//nf/6nX79+S5YsoUpiY2Nnz54dHh6uyxWa7pCTk3P69Ons7GztI5ZdtWfPnry8vAsXLrDZbITQmTNn3Nzcxo8fr/5oSJep7xaxHI7GxsZSA699+vQ5efIkw0s3XPcdjmJfM4b/Hn788cf169cbqz3YZWVl7dixQ/2JHD2gNoejBKl2Eyr1ijuyxz3t1q16/CsP4fdgXARBpKenU8/TUXrC4SgAZg2CEADMIAgBwAyCEADMIAgBwKydhDBtE1aATvXgldaDu2Yi2glC/V7R8dZKSkpCCGncBNwz3Lp1Kzk5GX4PxqXxkD5qNwjVRzBAp6gRwp660pKTk3tq13BpG4RwTggAZhCEAGAGQQgAZhCEAGAGQQgAZl0Owtu3bw8YMIDFYhEE4ezszMBbdNUTa7m4uGgk3wLYQWo0A1Oj6fk84ZQpUxBCdXV1hjxY1SVisdjGxoaxxenOBLMyGYvur7egU6O9fv2aLo+LiwsODqYzH1Gp0RBC58+fV/86rqxMEyZMaBsOU6ZMIUly+vTpu3fvrqqqamxszMjIYLPZ77//PvWt5OTkCRMm0L/81tZWOjWa2b/eQoNCoQgICMDdCvyMsh4YWJlWVlbUk/X0O3937dqVlpaWkZGh/rLzlJQUFoslkUhwPW6vjsfjtU2N9umnnyKEOBzOypUrHR0dhULh7NmzZ86cefnyZeoNbqtXrx42bNi0adNaWloQQgRBUE/WUy/G14+JBiH25F4mwijrgfmVCanRusQIQWgKac9++umngQMH2tjY8Hi8IUOG/Pjjjwihjz/+mDpYF4vF1BtmFy9ezOfzbWxszp49224+rX/84x98Pt/a2rqqqiomJsbNze3x48eGryLUcdIv3deDGWVKg9RoXaO+O9b7nJCBtGfazwlPnjy5ZcuWly9f1tbWjh07lj46Dw0NtbCwePHiBT3nvHnzzp49S3acT4vqy+rVq1NTUz/44IP/+7//074qdDwn1JL0S/f1wHCmNEiNZn6p0TCmPQsLC/v888/t7Ozs7e1DQkJqa2uprCnLly9XqVT04mQy2d27d6dNm9ZpPq1du3atWrXq9OnTfn5+hjdPl6RfOjL9TGmQGq2ruuWcEG9yL+pddNR15EmTJvXr1+/w4cPUf6C0tLTw8HALCwvD82l1SZeSfunONDOlQWq0rsJwYaY7knv98MMPEydOdHR05HK51AUuCkEQy5Yte/bs2dWrVxFC33333d/+9jdkjHxaXaJf0i9dmGCmNEiN1lVMB6Fxk3vduHEjKSmppKRk1qxZLi4ud+7caWhokEql6vMsWrSIx+MdPHjw8ePHIpGIOr02MJ9WV+mX9KtTppkpDVKjdRXTQWjc5F7/7//9P4FAcP/+faVSuWLFCh8fHx6Pp/EkuJ2d3Zw5c7Kysnbv3r106VKq0MB8Wl2lPemX3uvBNDOlQWq0rmIiCLsj7ZlSqaysrMzJyREIBJ6engihK1euvH79urCwsO0pxPLly9+8eXP+/Png4GCqREs+re6gPelXl9aD6WdKg9RoXaa+E9flkvTt27cHDRpEXUdycXHZvn17dyf3+vrrr7VcasvMzCRJcv369fb29ra2trNnz/7qq68QQmKxmLp2TxkxYkRsbKx6R9rNpyWVSqljDA8PDx2TCuk4RNFR0i/d10NFRQXDmdIgNVoPSY1mImnPpk2b9uzZs+6omcl7RxlemZAarV1mmRoNV3Iv+jg2Pz+f2kVgaYZxmWamNEiN9ranRuvI+vXrCwsLnzx5snjxYvoSHOgOkBrNdFOj4U3utXHjRhaL5eHhQd2n1k0YOxxlfmVCajQNkBrNREFqNNAlkBoNAJMDQQgAZhCEAGAGQQgAZu3kosjIyGC+HeaLuoOpR6406jboHtk106J+qRTy7wDAAG1DFMDsUFe6YWdl1uCcEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADMIAgBwAyCEADM2slZD0zZjRs3qFTylEePHiGEpFIpXeLv7/+Xv/wFQ8uAviBdtpm5evXqX//6VzabzWJpHsW0trYqlcorV65MnjwZS9uAfiAIzUxra6uLi0t1dXW7U3v16lVRUWFhYcFwq4Ah4JzQzLBYrPnz53M4nLaTOBxOREQERKDZgSA0P3Pnzm1ubm5b3tzcPHfuXObbAwwEh6NmqU+fPsXFxRqFHh4excXFBEFgaRLQG+wJzdKCBQvYbLZ6CZvNXrRoEUSgOYI9oVl69OjRgAEDNAofPHgwaNAgLO0BhoA9oVny8/MbNGiQ+n5v4MCBEIFmCoLQXH300Uf0hVA2m71w4UK87QF6g8NRc1VaWurl5UVtPoIgnj171qdPH9yNAvqAPaG58vDwePfdd1ksFovFevfddyECzRcEoRlbsGABQRAsFmvBggW42wL0B4ejZqympsbFxQUhVF5e7uTkhLs5QF+kYcLCwnD3AACcwsLCDAwiIzzKNHbs2DVr1hhej7m4detWcnJyeno67oYghNCNGzcIghg/fryxKpwzZ05UVJS/v7+xKuzZkpKSDK/ECEHo7u7+4YcfGl6PGUlOTjaRLk+dOhUhZG1tbawK58yZ4+/vbyK9M30nT540vBJ4qNe8GTH8AC5wdRQAzCAIAcAMghAAzCAIAcAMQxB+/PHH1tbWBEHk5eUxv/SOvH792s/Pb9OmTd1U/4ULF2xsbM6dO9dN9WNx5cqV2NjY06dP+/j4EARBEITGvTtBQUHW1tYWFhaDBg26d+8ek22bOHEi0YZQKEQISaVSPz8/KysrgUDg5+e3efNmmUyGEDp79qxUKlWpVEy2E2EJwoMHDx44cID55Wq3cePGx48fd1/9Pe/OpM8//zwlJWXDhg2hoaHPnj0Ti8UODg7Hjh374Ycf6HkuXbp08uTJ4ODggoKCkSNHYmwtJTAwECH0008/LV26tKSkpLKyctu2bVKplLrnJCQkhMfjTZ48ub6+nslWweEoQgj9/PPPDx486NZFTJ8+vaGhITg4uJvqVygUAQEB3VR5W7t27UpLS8vIyFAfI0lJSWGxWBKJpKGhgbGWdITH48lkMvUbUyQSyaeffooQ4nA4K1eudHR0FAqFs2fPnjlz5uXLl//44w+E0OrVq4cNGzZt2rSWlhbGmoonCE3qLQwKhWLdunXJycm4G2KQQ4cOVVVVMbOsp0+fbt68eevWrTweT708ICAgKirqxYsXa9euZaYlWly8eFH9H0RpaemDBw8mTZqEEMrMzFRvuZubG0KoqamJ+rhly5a8vDwmfw8MBSFJkomJif379+dyuTY2NuvWraMnqVSquLg4T09PKyuroUOHUreD7du3TyAQ8Pn8M2fOTJ06VSQSubu7nzhxgvrK9evXx4wZw+fzRSLRkCFDqAP6duvRxcaNG6n/i8bu9P/v5s2bnp6eBEF89dVXSGvvUlJSeDyek5PTsmXLXF1deTxeQEDAnTt3EEKRkZEcDoe6YxshtHLlSoFAQBBETU1NVFRUTExMUVERQRC+vr4IoYsXL4pEou3bt3dHd1JSUkiSDAkJaTspISGhX79+Bw8evHLlStupJEnu2bNnwIABXC7Xzs5u5syZ1BvEtW9uvbesul27dq1evbrdSYWFhba2tl5eXtRHOzu7CRMmJCcnM3cGYeC9p2FhYbrcwLpx40aCIL744ou6ujq5XL53716EUG5uLkmSa9eu5XK5p06dqqur27BhA4vFunv3LvUVhNDVq1cbGhqqqqrGjx8vEAiam5ubmppEIpFUKlUoFBUVFR988EF1dbWWerS7efNmSEgISZLU63Q3btzY6VeoH0Gns2koLS1FCKWmptIrpN3ekSQpkUgEAsHDhw9fv35dUFAwevRoa2vrkpISkiTnz5/v7OxM15mYmIgQorofGhoqFovpSefPn7e2to6Pj+9qOxFC6enp2ufx8fEZOHCgRqFYLP79999Jkvz5559ZLFafPn2amppIkszOzp4xYwY1T1xcHIfDOXr0aH19fX5+/siRI6m3FWtfIfptWXVlZWUDBw5UqVTqhc3NzWVlZampqVwu9+jRo+qTYmNj6d+ndjr+/rVjIgjlcjmfz3///ffpEuqfXG5urkKh4PP54eHh9JxcLnfFihXkn1tFoVBQk6i4ffr0KXXydv78efVFaKlHe8NGjRpVVlZGYgrCtr0jSVIikdjY2NBfvHv3LkJo69atZFeCUG+dBmFTUxNBEMHBwRrldBCSJBkTE4MQWrVqFakWhHK5XCgU0tuIJMlffvkFIUT9p+hohei3ZTWsWrXq66+/1ih0dnZGCDk4OHz55ZdUtNMOHz6MEPruu+86rdkoQcjE4ejTp0/lcnm7CRIeP34sl8sHDx5MfbSysnJxcaEOUTRQ75xWKpU+Pj5OTk4RERFbtmx5/vx5V+tRt2HDhr///e/UKQFedO/aTho1ahSfz++0L4ypqqoiSZLP52uZJyEhoX///nv37r158yZdWFBQ0NTUNGrUKLpk9OjRHA6HOtjWQK8Q/basuvLy8rNnzy5atEijvLS0tKqq6vjx499+++2IESPUz6ip3lVWVuq+FEMwEYRlZWUIoXZPul69eoUQ2rRpEz2SU1xcLJfLtdRmZWV17dq1wMDA7du3+/j4hIeHKxQKPeq5efPm/fv3P/74Y4P6xggul9tR8gnmvX79GiHE5XK1zMPj8Y4cOUIQxJIlSxQKBVVIXfenRupotra2jY2NWqrSY8tqkEqlS5cu1biGhBBis9mOjo5BQUFpaWkFBQU7duygJ1lZWdE9ZQATQUj1/82bN20nUZGZlJSkvndWT/3VrkGDBp07d668vHz9+vXp6em7d+/Wo55Dhw5dvXqVxWJRm5aqYfv27QRB/Prrr3p31uiUSmV9fb27uzvuhvwH9QPtdETb398/Ojq6sLBw27ZtVImtrS1CSCPkOu2afr8QWkVFxfHjx1esWKFlHl9fXwsLi4KCArqEyjJA9ZQBTATh4MGDWSzW9evX207y8PDg8XhdunWmvLz84cOHCCFHR8edO3eOHDny4cOHetRz5MgR9e2qfk6ofsiEXU5ODkmSY8eORQhZWlq2e8jKJCcnJ4IgdBkJ3LZtm5+fX25uLvVx8ODBQqFQ/R/cnTt3mpub33nnHS2V6LFl1Uml0oiICHt7e7qktrZ23rx56vMUFhaqVCoPDw+6hOodddLIACaC0NHRMSws7NSpU4cOHZLJZPn5+fv376cm8Xi8xYsXnzhxYt++fTKZTKVSlZWVUcOmHSkvL1+2bNmjR4+am5tzc3OLi4vHjh2rRz2mrLW1ta6urqWlJT8/PyoqytPTkzql8fX1ffnyZVZWllKprK6uVk9HYW9vX15e/vz588bGRqVSmZ2d3U1DFHw+38fHhzrF0I46KKVfjsrj8WJiYjIzM48dOyaTye7fv798+XJXV1eJRKK9ko62bHh4uLOzs5a74SorKw8fPqzx2geBQHDp0qVr167JZDKlUpmbm7tw4UKBQBAdHU3PQ/VuyJAhnfbROAy8sKPj1aHGxsalS5c6ODgIhcLAwMC4uDiEkLu7+2+//fbmzZv169d7enpaWlo6OjqGhoYWFBTs3buXOjnu27dvUVHR/v37RSIRQsjLy+vy5csBAQF2dnYWFha9e/feuHFjS0sLSZLt1qN7R7r16mhqaio1vsfn80NCQrT07smTJxKJhM1mu7m5WVpaikSimTNnFhUVUfXU1ta+9957PB7P29v7k08+oYZbfX19S0pK7t275+XlZWVlFRgYWFFRceHCBWtr64SEhC61k9RtiCIyMpLNZsvlcupjZmamWCxGCPXq1Yu6Iqpu3bp19BBFa2trYmJi37592Wy2nZ3drFmzHj9+TJKk9hXS0ZadNWsWQiguLq6jdkZHR0dERLQtDwkJ8fb2FgqFXC5XLBaHh4ffv39ffYbp06e7ubm1trZ2srLMaIiih9FviEJ3EonE3t6+++rXTpcgLCwstLS01BhbY55KpRo/fvyhQ4eMW21NTQ2Px9u9e7cuM5vNEAXoKuZv5O8SX1/f+Pj4+Ph4+lYv5qlUqqysrMbGxvDwcOPWvGXLluHDh0dGRhq3Wi16chA+evSo7cMsNKNvvLdKbGzs7Nmzw8PDcd2rnZOTc/r06ezsbO0jll21Z8+evLy8CxcuaGSe61Y9OQj9/Py0HAOkpaXhbmA7NmzYcOTIkYaGBm9v71OnTuFujjbbt2+PjIzcuXMnlqVPnjz5+++/p++kNYozZ868efMmJyfHzs7OiNV2Ct62Zlp27NihPmps4oKCgoKCgnC3wmhmzJgxY8YM5pfbk/eEAJgFCEIAMIMgBAAzCEIAMDPChZmysrKMjAzD6zEX1N3DPbjLut8eDcrKyoxwb72Bg/2QGg285Uzijhm4ba0nQTrctgZoRtkJwTkhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGbYgVM+nReFwOE5OThMnTkxMTKyrq8PVMKA7U06N1lH+M9Rx1jRcqdEwv95CLBZTb5umXm3073//e9GiRQRBuLq6dvVV54yBcUJKXFxccHAwnfmISo2G2rwcXf01+EyaPn367t27q6qqGhsbMzIy2Gw2/Q74CRMmtA2EKVOmkCSZnJw8YcKEuro6HZfSo15vQRCEra3txIkTjxw5kpGRUVlZSeUSw90uphklwxkDadJMPzWalvxnWrKmvUWp0bQLCwtbtGhRVVXVN998g7stTDNKhrPuTpNmFqnRtOQ/05I1DfXg1GhdRb1mMzs7G5lA7jT9kB2kAdM9w5nJpkkzx9RoGvnP1GlkTeuxqdE6Qp8TaqDCxsPDg8SdO60tHc8JtaQB0z25EvNp0lDPSo2mJf8Zpd2saT0wNZoWHQUhSZLUWSLe3Gnt0iUItacB61IQMpwmrdMgNK/UaFryn1HazZrWA1Oj6eHVq1ckSYpEIry50/TWpTRgujOFNGnmlRpNS/4z1HHWtB6YGk0PT548QQj5+flhzJ1mCP3SgOkCe5o080qNpiX/Geo4a1oPTI2mh4sXLyKEpk6dijF3miH0SwPWKVNIk2ZeqdFobfOfacma1gNTo3VVRUVFUlKSu7v7kiVLMOZOM4T2NGB6ZzgzhTRpZpEaTZf8Z22zptF6YGo07UiSbGpqojLgVFdXp6enjxs3zsLCIisrS/jWvNMAACAASURBVCQSmWnuNO1pwHTPcIZML02aWaRG6zT/WbtZ02g9MzVaW2fPnh06dCifz+dwOCwWC/1508yYMWPi4+Nra2vpOU0kdxpNxyGKjtKAkV3JcMZ8mjTUU1Kjac9/1lHWNAqkRjN1TN47ynyaNF2CEFKj0XryEAWgmWCaNEiNZlwQhEAfkBrNiCAITZeJp0mD1GjGAqnRTJfpp0mD1GhGAXtCADCDIAQAMwhCADCDIAQAMyNcmLl9+/bs2bMNr8dcUPc09eAuJyUlnTx5EncrzMPt27epW3kNQZCGPcO/Z88eSGeH0f379xGTdzmCNqjnRQypwdAgBHh9+OGHqEdnLH0bwDkhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBpl4z89133+3Zs0elUlEfa2pqEEK9evWiPlpYWERHR3/00UfY2ge6DoLQzDx58qR///5aZnj8+HG/fv0Yaw8wHByOmpl+/foNGzaMIIi2kwiCGDZsGESg2YEgND8fffSRhYVF23JLS8uFCxcy3x5gIDgcNT/l5eUeHh6tra0a5QRBlJaWurm5YWkV0BvsCc1P7969AwICWKz/2nYsFmvcuHEQgeYIgtAsLViwQKOEIAi4KGqm4HDULNXV1Tk7OyuVSrrE0tKyoqLCwcEBY6uAfmBPaJbs7Ozef/99+vKMhYXFlClTIALNFAShuYqIiKCvzZAkGRERgbc9QG9wOGqu5HK5g4PD69evEUI8Hq+mpkYgEOBuFNAH7AnNFZ/PnzVrFpvNZrPZs2bNggg0XxCEZmzevHlKpVKpVM6bNw93W4D+LA38/q1bt0pLS43SFNBVKpWKz+eTJCmTyTIyMnA35y3l4eHh7+9vUBWkYcLCwozUFwDMUlhYmIFBZITDUcMbYV7S09ORwf+8jCUnJ+f69etGrBAhlJ6ebsQKezaj7IQMPRwFeI0fPx53E4ChIAjNm8YdpMAcwSYEADMIQgAwgyAEADMIQgAwwxCEH3/8sbW1NUEQeXl5zC9dXUJCAvHfBg8e3E3LunDhgo2Nzblz57qpfiyuXLkSGxt7+vRpHx8fagVqPOgYFBRkbW1tYWExaNCge/fuMdk2qVTq5+dnZWUlEAj8/Pw2b94sk8moSRMnTiTaEAqFZ8+elUql9JvsGIMhCA8ePHjgwAHml4sX2eNulP/8889TUlI2bNgQGhr67NkzsVjs4OBw7NixH374gZ7n0qVLJ0+eDA4OLigoGDlyJJPN++mnn5YuXVpSUlJZWblt2zapVKp9TC8wMDAkJITH402ePLm+vp6xdiI4HD169Kj62OuDBw+6aUHTp09vaGgIDg7upvoVCkVAQEA3Vd7Wrl270tLSMjIyrK2t6cKUlBQWiyWRSBoaGhhrSUc4HM7KlSsdHR2FQuHs2bNnzpx5+fLlP/74AyHE4/FkMpn6dpdIJJ9++ilCaPXq1cOGDZs2bVpLSwtjTcUThO2+sQ8Y4tChQ1VVVcws6+nTp5s3b966dSuPx1MvDwgIiIqKevHixdq1a5lpiRaZmZnqzaPevtPU1IQQunjxovr/jtLS0gcPHkyaNIn6uGXLlry8vOTkZMaaylAQkiSZmJjYv39/LpdrY2Ozbt06epJKpYqLi/P09LSysho6dCh1U9i+ffsEAgGfzz9z5szUqVNFIpG7u/uJEyeor1y/fn3MmDF8Pl8kEg0ZMoQ61m+3HhNx8+ZNT09PgiC++uorpLV3KSkpPB7Pyclp2bJlrq6uPB4vICDgzp07CKHIyEgOh+Pi4kLVuXLlSoFAQBBETU1NVFRUTExMUVERQRC+vr4IoYsXL4pEou3bt3dHd1JSUkiSDAkJaTspISGhX79+Bw8evHLlStupJEnu2bNnwIABXC7Xzs5u5syZjx490r5CkJG2bGFhoa2trZeXV9tJu3btWr16Nf3Rzs5uwoQJycnJzJ1BGH7vnC73jm7cuJEgiC+++KKurk4ul+/duxchlJubS5Lk2rVruVzuqVOn6urqNmzYwGKx7t69S30FIXT16tWGhoaqqqrx48cLBILm5uampiaRSCSVShUKRUVFxQcffFBdXa2lHi22bdvm7u5ua2vLZrP79OkzY8aMX375pdO+6HfvKPWsSWpqKr1C2u0dSZISiUQgEDx8+PD169cFBQWjR4+2trYuKSkhSXL+/PnOzs50nYmJiQghqvuhoaFisZiedP78eWtr6/j4+K62E+lw76iPj8/AgQM1CsVi8e+//06S5M8//8xisfr06dPU1ESSZHZ29owZM6h54uLiOBzO0aNH6+vr8/PzR44c2atXr4qKCu0rRI8tS2tubi4rK0tNTeVyuRqnHpSysrKBAweqVCr1wtjYWPr3qZ2Ov3/tmAhCuVzO5/Pff/99uoT6J5ebm6tQKPh8fnh4OD0nl8tdsWIF+edWUSgU1CQqbp8+fUqdtp0/f159EVrq0aKkpOTevXuNjY1v3ry5devWiBEjrKysHjx4oP1bRgzCtr0jSVIikdjY2NBfvHv3LkJo69atZFeCUG+dBmFTUxNBEMHBwRrldBCSJBkTE4MQWrVqFakWhHK5XCgU0tuIJMlffvkFIUT9p+hohei3ZWnOzs4IIQcHhy+//JIKaQ2rVq36+uuvNQoPHz6MEPruu+86rd8oQcjE4ejTp0/lcvnkyZPbTnr8+LFcLqcHBqysrFxcXKhDFA0cDgchpFQqfXx8nJycIiIitmzZ8vz5867Wo87Dw2PEiBFCoZDD4YwdO/bIkSMKhYLa/Ayje9d20qhRo/h8fqd9YUxVVRVJknw+X8s8CQkJ/fv337t3782bN+nCgoKCpqamUaNG0SWjR4/mcDjUwbYGeoXot2VppaWlVVVVx48f//bbb0eMGKFx2lxeXn727NlFixZpfIvqXWVlpY5LMRATQVhWVoYQcnR0bDvp1atXCKFNmzbRwzXFxcVyuVxLbVZWVteuXQsMDNy+fbuPj094eLhCodCjnraGDBliYWHx5MmTLn2LAVwut7q6Gncr/oN6qw2Xy9UyD4/HO3LkCEEQS5YsUSgUVCF13V8oFKrPaWtr29jYqKUqA7csm812dHQMCgpKS0srKCjYsWOH+lSpVLp06VKNy0sIISsrK7qnDGAiCKlOvnnzpu0kKjKTkpLU9863bt3SXuGgQYPOnTtXXl6+fv369PT03bt361ePhtbW1tbWVu0/L+Yplcr6+np3d3fcDfkP6gfa6Yi2v79/dHR0YWHhtm3bqBJbW1uEkEbIddo1o2xZhJCvr6+FhUVBQQFdUlFRcfz48RUrVrSdubm5Gf3ZUwYwEYSDBw9msVjXr19vO8nDw4PH43Xp1pny8vKHDx8ihBwdHXfu3Dly5MiHDx/qUQ9CaMqUKeofqdN9Q19VYGw5OTkkSY4dOxYhZGlp2e4hK5OcnJwIgtBlJHDbtm1+fn65ubnUx8GDBwuFwl9//ZWe4c6dO83Nze+8846WSvTbsrW1tRrv3SksLFSpVB4eHnSJVCqNiIiwt7dv+3Wqd9T5JAOYCEJHR8ewsLBTp04dOnRIJpPl5+fv37+fmsTj8RYvXnzixIl9+/bJZDKVSlVWVkaNqHakvLx82bJljx49am5uzs3NLS4uHjt2rB71IIRevHiRlpZWX1+vVCpv3br18ccfe3p6Ll++3Gg911dra2tdXV1LS0t+fn5UVJSnpyd13uLr6/vy5cusrCylUlldXV1cXEx/xd7evry8/Pnz542NjUqlMjs7u5uGKPh8vo+PD3WKoR11UEq/oZjH48XExGRmZh47dkwmk92/f3/58uWurq4SiUR7JR1t2fDwcGdn53bvhhMIBJcuXbp27ZpMJlMqlbm5uQsXLhQIBNHR0dQMlZWVhw8fXrNmTbsLpXo3ZMiQTvtoHAZe2NHx6lBjY+PSpUsdHByEQmFgYGBcXBxCyN3d/bfffnvz5s369es9PT0tLS0dHR1DQ0MLCgr27t1LnRz37du3qKho//79IpEIIeTl5XX58uWAgAA7OzsLC4vevXtv3LixpaWFJMl269HeqpiYGLFYLBAILC0t3d3dly5dWl5e3mlf9Lg6mpqaSo3v8fn8kJAQLb178uSJRCJhs9lubm6WlpYikWjmzJlFRUVUPbW1te+99x6Px/P29v7kk0+o4VZfX1/qMq+Xl5eVlVVgYGBFRcWFCxesra0TEhK61E5StyGKyMhINpstl8upj5mZmWKxGCHUq1cv6oqounXr1tFDFK2trYmJiX379mWz2XZ2drNmzXr8+DFJktpXSEdbdtasWQihuLi4dhsZEhLi7e0tFAq5XK5YLA4PD79//z49NTo6OiIioqMOTp8+3c3NrbW1tZOVZUZDFD1Md79jRiKR2Nvbd1/92ukShIWFhZaWlu0OuzFJpVKNHz/+0KFDxq22pqaGx+Pt3r1bl5nNZogCdBXzN/J3ia+vb3x8fHx8PHUXGBYqlSorK6uxsTE8PNy4NW/ZsmX48OGRkZHGrVaLnhyEjx49avvECs3oG++tEhsbO3v27PDwcFz3aufk5Jw+fTo7O1v7iGVX7dmzJy8v78KFC2w224jVateTg9DPz0/LMUBaWhruBrZjw4YNR44caWho8Pb2PnXqFO7maLN9+/bIyMidO3diWfrkyZO///57+k5aozhz5sybN29ycnLs7OyMWG2n4G1rpmXHjh0aA8qmLCgoKCgoCHcrjGbGjBkzZsxgfrk9eU8IgFmAIAQAMwhCADCDIAQAMyNcmLl9+/bs2bMNr8dcUPc09eAuJyUlnTx5EncrzMPt27ep23oNAXtCADAzwp5w7Nixb9U/zoyMjDlz5vTULhMEsWbNmg8//BB3Q8yDUQ6IYE8IAGYQhABgBkEIAGYQhABgBkEIAGbYglA9lQ+Fw+E4OTlNnDgxMTGxrq4OV8OA7kw5KxOltbU1KSmpbZaO48ePU29V9vLyWrx4cUVFBUIIV1YmzE/Wi8Vi6kW31FtV/v3vfy9atIggCFdXV93fssyw7n6yHi+kw5P1lLi4uODgYDqzCpWVCbV5L7P6G7gZ9uTJk3HjxiGEhg0bpl5OPcUmlUrr6+tzc3N9fHyGDx+uVCpJkkxOTp4wYUJdXZ2Oi+hRT9YTBGFraztx4sQjR45kZGRUVlZSaYxwt4tpRkmuxECGJtPPyvTbb7999tlny5cvHz58uMakf/7zn7179163bp2Njc3w4cOjo6Pz8vKodxC/RVmZtAsLC1u0aFFVVdU333yDuy1MM0pype7O0GQWWZmGDRt2+vTp+fPnt32RbGlpqaurK50ajHoPIv3quh6blamrqDf8ZWdnI7NN20R2kIFI9+RKJpuhyRyzMqnz8fFR/ydFnRD6+PhQH3tsVqaO0OeEGqiw8fDwILGmbWqXjueEWjIQ6Z7XhfkMTahnZWUiSfLdd9/VOCfMyclhs9kpKSkymezBgwcDBgyYMmWK+gw9MCuTFh0FIUmS1Fki3rRN7dIlCLVnIOpSEDKcoanTIDSvrExke0FIkuSmTZvo/ZC7u3tpaan61B6YlUkPr169IklSJBLhTdukty5lINKdKWRoMq+sTO3auHHj/v37r1692tTU9OzZs4CAAH9/fyp3HaUHZmXSA5Uayc/Pz6TSNulOvwxEusCeocm8sjK19ccff0il0r///e+TJk0SCATe3t4HDhwoLy+nDiUoPTArkx4uXryIEJo6dapJpW3SnX4ZiDplChmazDQrE43KDNO7d2+6RCQS2dvbqyds6oFZmbqqoqIiKSnJ3d19yZIleNM26U17BiK9kyuZQoYms8jKpAUV8+rJghobG1++fKmesKkHZmXSjiTJpqYmKvlGdXV1enr6uHHjLCwssrKyRCIR3rRNetOegUj35ErI9DI0mUVWJi28vb3fe++9AwcO3LhxQ6FQlJaWUg3429/+Rs/TM7MytXX27NmhQ4fy+XwOh8NisdCfN82MGTMmPj6+traWnhNj2qZ26ThE0VEGIrIryZWYz9CEekpWplu3bo0bN87V1ZX6nbu4uAQEBFy/fp0kSWo01dfXl8vlCoXCcePG/etf/1L/LmRlMnVM3jvKfIYmXYIQsjLRevIQBaCZYIYmyMpkXBCEQB+QlcmIIAhNl4lnaIKsTMYCWZlMl+lnaIKsTEYBe0IAMIMgBAAzCEIAMIMgBAAzCEIAcDNwsD8sLAx3DwDAyfA7ZgjSsBdp3Lp1S/1pSMCwpKQkhNCaNWtwN+Tt5eHh4e/vb0gNhgYhwIvKYZaRkYG7IUB/cE4IAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgBkEIAGYQhABgZom7AaBrampqZDIZ/fHVq1cIoWfPntElIpGoV69eGFoG9AWZes3MkSNHlixZomWGw4cPL168mLH2AMNBEJqZhoYGR0dHpVLZ7lQ2m11dXW1jY8Nwq4Ah4JzQzNjY2EybNs3Ssp3zCEtLy+nTp0MEmh0IQvMTERGhUqnalre2tkZERDDfHmAgOBw1P69fv+7Vqxd1SUYdn8+vqamxsrLC0iqgN9gTmh8ej/fBBx+w2Wz1QjabHRYWBhFojiAIzdK8efM0rs0olcp58+bhag8wBByOmqWWlhZnZ+eXL1/SJba2ttXV1e1esAEmDvaEZsnS0nLu3Ln0ESmbzY6IiIAINFMQhOZq7ty59BGpUqmcO3cu3vYAvcHhqLkiSdLDw+PFixcIIVdX1xcvXhAEgbtRQB+wJzRXBEEsWLCAw+FwOJyFCxdCBJov2BOasfz8/GHDhlF/DBkyBHdzgJ4MPZXfs2fPrVu3jNIUoAehUIgQio+Px92Qt5e/v390dLQhNRh6OHrr1q3bt28bWIl5KSsrO3XqFO5W/IeXl1efPn2MWOGpU6fKysqMWGHPdvv2bcN3Qka4qD127NiTJ08aXo+5yMjImDNnjol0mXqS0MfHx1gVEgSxZs2aDz/80FgV9myzZ882vBIYWTJvRgw/gAtcHQUAMwhCADCDIAQAMwhCADDDEIQff/yxtbU1QRB5eXnML12DUqncsWOHr68vh8OxtbUdPHjw8+fPu2NBFy5csLGxOXfuXHdUjsuVK1diY2NPnz7t4+NDEAR1E4/6DEFBQdbW1hYWFoMGDbp37x7zLWxtbU1KSgoICNAoP378+OjRo62trb28vBYvXlxRUYEQOnv2rFQqbfetBd0KQxAePHjwwIEDzC+3XXPmzPnuu+++//57uVz+f//3f2KxuKmpqTsW1PPuTPr8889TUlI2bNgQGhr67NkzsVjs4OBw7NixH374gZ7n0qVLJ0+eDA4OLigoGDlyJMMtLCws/Mtf/hIdHS2Xy9XL09PT58+fP3v27LKysjNnzty4cWPq1KktLS0hISE8Hm/y5Mn19fVMtvOtPhxNS0vLyso6efLku+++a2lp6erqeubMmcGDB3fHsqZPn97Q0BAcHNwdlSOEFApF2//33WfXrl1paWkZGRnW1tZ0YUpKCovFkkgkDQ0NjLWkI7/99ttnn322fPny4cOHa0z65z//2bt373Xr1tnY2AwfPjw6OjovL+/OnTsIodWrVw8bNmzatGktLS2MNRVPEJrI3cZff/31yJEje8Zdl4cOHaqqqmJmWU+fPt28efPWrVt5PJ56eUBAQFRU1IsXL9auXctMS7QYNmzY6dOn58+fz+VyNSaVlpa6urrSP0IPDw+EUHFxMfVxy5YteXl5ycnJjDWVoSAkSTIxMbF///5cLtfGxmbdunX0JJVKFRcX5+npaWVlNXTo0PT0dITQvn37BAIBn88/c+bM1KlTRSKRu7v7iRMnqK9cv359zJgxfD5fJBINGTKEeiN1u/Vo0dzcfPv27bb/JrvDzZs3PT09CYL46quvkNbepaSk8Hg8JyenZcuWubq68ni8gIAA6p90ZGQkh8NxcXGh6ly5cqVAICAIoqamJioqKiYmpqioiCAIX19fhNDFixdFItH27du7ozspKSkkSYaEhLSdlJCQ0K9fv4MHD165cqXtVJIk9+zZM2DAAC6Xa2dnN3PmzEePHmlfIajrW7ZTPj4+6v+wqBNC+rYHOzu7CRMmJCcnM3cGQRomLCwsLCys09k2btxIEMQXX3xRV1cnl8v37t2LEMrNzSVJcu3atVwu99SpU3V1dRs2bGCxWHfv3qW+ghC6evVqQ0NDVVXV+PHjBQJBc3NzU1OTSCSSSqUKhaKiouKDDz6orq7WUk9Hfv/9d4TQ8OHDJ06c6OLiwuVy/fz8vvrqq9bWVu19oX4EXVhHJEmSZGlpKUIoNTWVXiHt9o4kSYlEIhAIHj58+Pr164KCAur6QUlJCUmS8+fPd3Z2putMTExECFHdDw0NFYvF9KTz589bW1vHx8d3tZ0IofT0dO3z+Pj4DBw4UKNQLBb//vvvJEn+/PPPLBarT58+TU1NJElmZ2fPmDGDmicuLo7D4Rw9erS+vj4/P3/kyJG9evWqqKjQvkK6umU1vPvuu8OGDVMvycnJYbPZKSkpMpnswYMHAwYMmDJlivoMsbGx9O9TOx1//9oxEYRyuZzP57///vt0CfVPLjc3V6FQ8Pn88PBwek4ul7tixQryz62iUCioSVTcPn369MGDBwih8+fPqy9CSz0duX//PkLo/fff/9///d/a2tr6+vrPPvsMIXTs2DHt3TFiELbtHUmSEonExsaG/uLdu3cRQlu3biW7EoR66zQIm5qaCIIIDg7WKKeDkCTJmJgYhNCqVatItSCUy+VCoZDeRiRJ/vLLLwgh6j9FRytEjy2roW0QkiS5adMmej/k7u5eWlqqPvXw4cMIoe+++67Tyo0ShEwcjj59+lQul0+ePLntpMePH8vlcvpaiJWVlYuLC3WIooHD4SCElEqlj4+Pk5NTRETEli1b6OEE3euhUacKgwYNCggIsLe3t7Gx2bp1q42Nzf79+/Xvqr7o3rWdNGrUKD6fr70vTKqqqiJJks/na5knISGhf//+e/fuvXnzJl1YUFDQ1NQ0atQoumT06NEcDoc62NZArxA9tmynNm7cuH///qtXrzY1NT179iwgIMDf35/6L0mheldZWWnIUnTHRBBSj8Y4Ojq2nUS9wXbTpk3En4qLizUuKGuwsrK6du1aYGDg9u3bfXx8wsPDFQqFHvW4uroihGpqaugSDofj5eVVVFSkVy+7EZfLra6uxt2K/3j9+jX6819YR3g83pEjRwiCWLJkiUKhoAqp6/7UA5A0W1vbxsZGLVXpsWW1++OPP6RS6d///vdJkyYJBAJvb+8DBw6Ul5dThxUU6vWtVE8ZwEQQUtfQ3rx503YSFZlJSUnqe+dOH9AaNGjQuXPnysvL169fn56evnv3bj3qEQqFffv2ffjwoXphS0uLqeVyUCqV9fX17u7uuBvyH9QPtNMRbepR18LCwm3btlEltra2CCGNkOu0a/r9QrQoLCxUqVS9e/emS0Qikb29fUFBAV3S3NyM/uwpA5gIwsGDB7NYrOvXr7ed5OHhwePxunTrTHl5ORU5jo6OO3fuHDly5MOHD/WoByE0Z86c3NxcOrmfXC4vLi42tRGLnJwckiTHjh2LELK0tOwoHxNjnJycCILQZSRw27Ztfn5+ubm51MfBgwcLhcJff/2VnuHOnTvNzc3vvPOOlkr027JaUDH/xx9/0CWNjY0vX76kBiooVO+cnZ2NtVDtmAhCR0fHsLCwU6dOHTp0SCaT5efn0+ddPB5v8eLFJ06c2Ldvn0wmU6lUZWVl6iuorfLy8mXLlj169Ki5uTk3N7e4uHjs2LF61IMQio6O9vLyWrRoUUlJSW1t7fr16xUKBXV5Bq/W1ta6urqWlpb8/PyoqChPT89FixYhhHx9fV++fJmVlaVUKqurq+mhLYSQvb19eXn58+fPGxsblUpldnZ2Nw1R8Pl8Hx8fXZ6+pw5KLSws6I8xMTGZmZnHjh2TyWT3799fvny5q6urRCLRXklHWzY8PNzZ2bmrd8N5e3u/9957Bw4cuHHjhkKhKC0tpRrwt7/9jZ6H6h1z/44NvLCj49WhxsbGpUuXOjg4CIXCwMDAuLg4hJC7u/tvv/325s2b9evXe3p6WlpaOjo6hoaGFhQU7N27lzo57tu3b1FR0f79+0UiEULIy8vr8uXLAQEBdnZ2FhYWvXv33rhxY0tLC0mS7dbTacNKS0vnzp1rZ2fH5XLHjBmTnZ3d6Vf0uDqamppKje/x+fyQkBAtvXvy5IlEImGz2W5ubpaWliKRaObMmUVFRVQ9tbW17733Ho/H8/b2/uSTT6jhVl9f35KSknv37nl5eVlZWQUGBlZUVFy4cMHa2johIaFL7SR1G6KIjIxks9lyuZz6mJmZKRaLEUK9evWiroiqW7duHT1E0drampiY2LdvXzabbWdnN2vWrMePH5MkqX2FdLRlZ82ahRCKi4trt5G3bt0aN24cdeaPEHJxcQkICLh+/TpJktTIqq+vL5fLFQqF48aN+9e//qX+3enTp7u5uXU6WEWa0RBFD6PfEIXuJBKJvb1999WvnS5BWFhYaGlpefToUWaa1BGVSjV+/PhDhw4Zt9qamhoej7d7925dZjabIQrQVczfyN8lvr6+8fHx8fHx3XSzuy5UKlVWVlZjY2N4eLhxa96yZcvw4cMjIyONW60WPTkIHz16x2FyIgAAFZVJREFURHTM6BvvrRIbGzt79uzw8HBc92rn5OScPn06Oztb+4hlV+3ZsycvL+/ChQsamee6VU8OQj8/Py3HAGlpabgb2I4NGzYcOXKkoaHB29vbdF6s2K7t27dHRkbu3LkTy9InT578/fff03fSGsWZM2fevHmTk5NjZ2dnxGo7BW9bMy07duzYsWMH7lboKigoKCgoCHcrjGbGjBkzZsxgfrk9eU8IgFmAIAQAMwhCADCDIAQAMwhCAHAzcLA/LCwMdw8AwMnwO2aMk5VpzZo1htdjLm7dupWcnGz4m05M05w5c6Kiovz9/XE3xDwkJSUZXokRgtDd3f1ty6SVnJzcU7s8Z84cf3//nto7ozNKhjw4JwQAMwhCADCDIAQAMwhCADCDIAQAM2xBqJ5Pi8LhcJycnCZOnJiYmFhXV4erYUB3ppwaTSqV+vn5WVlZCQQCPz+/zZs3U+kSEEITJ05s+3ypUCjElRoN8+stxGIx9bZp6tVG//73vxctWkQQhKura5dedc6k7n69BV5Ih9dbUOLi4oKDg2UyGfWRSo2G2rwcXf01+EyaPn367t27q6qqGhsbMzIy2Gw2/Q74CRMmtA0E6k34ycnJEyZMqKur03EpPer1FgRB2NraTpw48ciRIxkZGZWVlVQuMdztYppRMpwxkCbN9FOjcTiclStXOjo6CoXC2bNnz5w58/Lly9Rr2ng8Hv2/gyKRSD799FP0VqVG0y4sLGzRokVVVVXffPMN7rYwzSgZzro7TZpZpEbLzMxUb56bmxtCiHopzsWLF9X/d5SWlj548GDSpEnUxx6bGq2rqNdsZmdnI3y50wxEdpAGTPcMZyabJs0cU6MVFhba2tp6eXm1nbRr167Vq1fTH3tsarSO0OeEGqiw8fDwIPHlTuuIjueEWtKA6Z5cifk0aahnpUZrbm4uKytLTU3lcrntvqOxrKxs4MCBKpVKvbAHpkbToqMgJEmSOkvEmDutI7oEofY0YF0KQobTpHUahOaVGo16lb2Dg8OXX35JhbSGVatWff311xqFPTA1mh5evXpFkqRIJMKYO80QXUoDpjtTSJNmXqnRSktLq6qqjh8//u23344YMULjVLm8vPzs2bPUuY+6HpgaTQ9PnjxBCPn5+WHMnWYI/dKA6QJ7mjTzSo3GZrMdHR2DgoLS0tIKCgo03mQnlUqXLl2qcXkJ9cjUaHq4ePEiQmjq1KkYc6cZQr80YJ0yhTRpZpoazdfX18LCQj3/WUVFxfHjx1esWNF25h6YGq2rKioqkpKS3N3dlyxZgjd3mt60pwHTO8OZKaRJM4vUaLW1tfPmzVMvodISquc/k0qlERER9vb2bb/eA1OjaUeSZFNTE5UBp7q6Oj09fdy4cRYWFllZWSKRCG/uNL1pTwOme4YzZHpp0swiNZpAILh06dK1a9dkMplSqczNzV24cKFAIIiOjqZmqKysPHz4cEdvhOiZqdHaOnv27NChQ/l8PofDYbFY6M+bZsaMGRMfH19bW0vPiTd3Wls6DlF0lAaM7EqGM+bTpKGekhotJCTE29tbKBRyuVyxWBweHn7//n16anR0dEREREcdhNRopo7Je0eZT5OmSxBCajRaTx6iADQTTJMGqdGMC4IQ6ANSoxkRBKHpMvE0aZAazVggNZrpMv00aZAazShgTwgAZhCEAGAGQQgAZhCEAGBmhAszZWVlGRkZhtdjLqi7h3twl7vvvvaep6yszAj30xs42A+p0cBbzvA7ZgiSsRdpgG5ApU/qwbvltwGcEwKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAGQQhAJhBEAKAmRFy1gMm3bhxQz2n/KNHjxBCUqmULvH39//LX/6CoWVAX5Au28xcvXr1r3/9K5vNZrE0j2JaW1uVSuWVK1cmT56MpW1APxCEZqa1tdXFxaW6urrdqb169aqoqLCwsGC4VcAQcE5oZlgs1vz58zkcTttJHA4nIiICItDsQBCan7lz5zY3N7ctb25unjt3LvPtAQaCw1Gz1KdPn+LiYo1CDw+P4uJigiCwNAnoDfaEZmnBggVsNlu9hM1mL1q0CCLQHMGe0Cw9evRowIABGoUPHjwYNGgQlvYAQ8Ce0Cz5+fkNGjRIfb83cOBAiEAzBUForj766CP6QiibzV64cCHe9gC9weGouSotLfXy8qI2H0EQz54969OnD+5GAX3AntBceXh4vPvuuywWi8VivfvuuxCB5guC0IwtWLCAIAgWi7VgwQLcbQH6g8NRM1ZTU+Pi4oIQKi8vd3Jywt0coC/SMGFhYbh7AABOYWFhBgaRER5lGjt27Jo1awyvx1zcunUrOTk5PT0dd0MQQujGjRsEQYwfP95YFc6ZMycqKsrf399YFfZsSUlJhldihCB0d3f/8MMPDa/HjCQnJ5tIl6dOnYoQsra2NlaFc+bM8ff3N5Hemb6TJ08aXgk81GvejBh+ABe4OgoAZhCEAGAGQQgAZhCEAGCGIQg//vhja2trgiDy8vKYX7q6iRMnEm0IhcLuWNaFCxdsbGzOnTvXHZXjcuXKldjY2NOnT/v4+FBrT+PenaCgIGtrawsLi0GDBt27d4/JtkmlUj8/PysrK4FA4Ofnt3nzZplMRk3qaLufPXtWKpWqVCom24mwBOHBgwcPHDjA/HJ1FBgY2B3V9rw7kz7//POUlJQNGzaEhoY+e/ZMLBY7ODgcO3bshx9+oOe5dOnSyZMng4ODCwoKRo4cyWTzfvrpp6VLl5aUlFRWVm7btk0qlWq/sSQwMDAkJITH402ePLm+vp6xdqK3/HCUx+PJZDL1exckEsmnn37aHcuaPn16Q0NDcHBwd1SOEFIoFAEBAd1UeVu7du1KS0vLyMhQHyNJSUlhsVgSiaShoYGxlnSEw+GsXLnS0dFRKBTOnj175syZly9f/uOPP5DW7b569ephw4ZNmzatpaWFsabiCUITeQvDxYsX1X9DpaWlDx48mDRpEsYm6e3QoUNVVVXMLOvp06ebN2/eunUrj8dTLw8ICIiKinrx4sXatWuZaYkWmZmZ6s1zc3NDCDU1NaHOtvuWLVvy8vKSk5MZaypDQUiSZGJiYv/+/blcro2Nzbp16+hJKpUqLi7O09PTyspq6NCh1O1g+/btEwgEfD7/zJkzU6dOFYlE7u7uJ06coL5y/fr1MWPG8Pl8kUg0ZMgQ6li/3Xq6ZNeuXatXrzZSj//LzZs3PT09CYL46quvkNbepaSk8Hg8JyenZcuWubq68ni8gICAO3fuIIQiIyM5HA51xzZCaOXKlQKBgCCImpqaqKiomJiYoqIigiB8fX0RQhcvXhSJRNu3b++O7qSkpJAkGRIS0nZSQkJCv379Dh48eOXKlbZTSZLcs2fPgAEDuFyunZ3dzJkzqTeIa9/chm9ZhFBhYaGtra2Xl1fbSRrb3c7ObsKECcnJycydQRh472lYWJguN7Bu3LiRIIgvvviirq5OLpfv3bsXIZSbm0uS5Nq1a7lc7qlTp+rq6jZs2MBise7evUt9BSF09erVhoaGqqqq8ePHCwSC5ubmpqYmkUgklUoVCkVFRcUHH3xQXV2tpR4dlZWVDRw4UKVSdTon9SPQvWZKaWkpQig1NZVeIe32jiRJiUQiEAgePnz4+vXrgoKC0aNHW1tbl5SUkCQ5f/58Z2dnus7ExESEENX90NBQsVhMTzp//ry1tXV8fHxX24kQSk9P1z6Pj4/PwIEDNQrFYvHvv/9OkuTPP//MYrH69OnT1NREkmR2dvaMGTOoeeLi4jgcztGjR+vr6/Pz80eOHEm9rVj7CjFkyzY3N5eVlaWmpnK53KNHj7adod3tHhsbS/8+tdPx968dE0Eol8v5fP77779Pl1D/5HJzcxUKBZ/PDw8Pp+fkcrkrVqwg/9wqCoWCmkTF7dOnTx88eIAQOn/+vPoitNSjo1WrVn399de6zGnEIGzbO5IkJRKJjY0N/cW7d+8ihLZu3Up2JQj11mkQNjU1EQQRHBysUU4HIUmSMTExCKFVq1aRakEol8uFQiG9jUiS/OWXXxBC1H+KjlaIgVvW2dkZIeTg4PDll19SIa2h3e1++PBhhNB3333Xaf1GCUImDkefPn0ql8vbTZDw+PFjuVw+ePBg6qOVlZWLiwt1iKKBeue0Uqn08fFxcnKKiIjYsmXL8+fPu1pPu8rLy8+ePbto0aKudcx46N61nTRq1Cg+n697X7pbVVUVSZJ8Pl/LPAkJCf3799+7d+/NmzfpwoKCgqamplGjRtElo0eP5nA41MG2BnqFGLhlS0tLq6qqjh8//u23344YMULjtLmj7U71rrKyUselGIiJICwrK0MIOTo6tp306tUrhNCmTZvo4Zri4mK5XK6lNisrq2vXrgUGBm7fvt3Hxyc8PFyhUOhRjzqpVLp06VKNywymg8vldpR8gnmvX79GCHG5XC3z8Hi8I0eOEASxZMkShUJBFVLX/TWGYW1tbRsbG7VUZeCWZbPZjo6OQUFBaWlpBQUFO3bsUJ/a0Xa3srKie8oAJoKQ6uSbN2/aTqIiMykpSX3vrJ76q12DBg06d+5ceXn5+vXr09PTd+/erV89lIqKiuPHj69YsaLLHWOEUqmsr693d3fH3ZD/oH6gnY5o+/v7R0dHFxYWbtu2jSqxtbVFCGmEXKddM2TLqvP19bWwsCgoKKBLtGx3KssA1VMGMBGEgwcPZrFY169fbzvJw8ODx+N16daZ8vLyhw8fIoQcHR137tw5cuTIhw8f6lEPTSqVRkRE2Nvb6/FdBuTk5JAkOXbsWISQpaVlu4esTHJyciIIQpeRwG3btvn5+eXm5lIfBw8eLBQKf/31V3qGO3fuNDc3v/POO1oq0W/L1tbWzps3T72ksLBQpVJ5eHjQJVq2O9U76nySAUwEoaOjY1hY2KlTpw4dOiSTyfLz8/fv309N4vF4ixcvPnHixL59+2QymUqlKisro0ZUO1JeXr5s2bJHjx41Nzfn5uYWFxePHTtWj3oolZWVhw8fNrU3A7S2ttbV1bW0tOTn50dFRXl6elLnLb6+vi9fvszKylIqldXV1erpKOzt7cvLy58/f97Y2KhUKrOzs7tpiILP5/v4+FCnGNpRB6X0y1F5PF5MTExmZuaxY8dkMtn9+/eXL1/u6uoqkUi0V9LRlg0PD3d2dm73bjiBQHDp0qVr167JZDKlUpmbm7tw4UKBQBAdHU3NoH27U70bMmRIp300DgMv7Oh4daixsXHp0qUODg5CoTAwMDAuLg4h5O7u/ttvv71582b9+vWenp6WlpaOjo6hoaEFBQV79+6lTo779u1bVFS0f/9+kUiEEPLy8rp8+XJAQICdnZ2FhUXv3r03btzY0tJCkmS79XTasOjo6IiIiC51WY+ro6mpqdT4Hp/PDwkJ0dK7J0+eSCQSNpvt5uZmaWkpEolmzpxZVPT/tXf/IOm8cQDHHyvNoiQjgqgI0cApmloqNAIjpERIcJSWptwaapCQ/gyKNraEU1D2B2vQthydImjpD98lHMwSCSto0PsNx++I1Ke787rnuevz2uIb55V98Hs+j/f+xx6nUChMTU3p9XqTybS8vMwut1oslsfHx6urq6Ghoba2tomJiVwul0wmOzs7NzY2BJ0nw2+Jwu/3a7Xaj48P9svT01Oz2YwQ6unpYd8R/WplZYVboqhUKqFQaHh4WKvVGo1Gt9t9d3fHMAz+F1LvmXW73QihQCBQ8yTn5+dNJlNHR0dra6vZbPZ6vTc3N9y/4p93p9PZ399fqVR++GUpaIlCZcQtUfC3tLTU3d39e8fH4zOEDw8PLS0tNZfd5FQulycnJ/f29qQ97MvLi16vD4fDfL5ZMUsUQCj5N/ILYrFYgsFgMBhkd4ERUS6XE4lEqVTyer3SHnl9fX10dNTv90t7WAw1D+Ht7W31J1Y4kj95f8rq6qrH4/F6vaT2aqfT6ZOTk1QqhV+xFCoSiVxfXyeTyW/luV+l5iG0Wq2Y/wMcHByQPsEa1tbWYrHY6+uryWQ6Pj4mfTo4m5ubfr9/e3ubyKNPT0/v7+9zO2klcXZ29vn5mU6njUajhIf9EdxtjS5bW1vfFpRp5nA4HA4H6bOQjMvlcrlc8j+uml8JAVAEGEIACIMhBIAwGEIACJPgjZlsNhuPxxs/jlKwu4dV/COL2B79Z2WzWQn21je42A9pNPDHUbFjBratqQnisW0NcCR5EYJrQgAIgyEEgDAYQgAIgyEEgDAYQgAIgyEEgDBiQ/i1p8XS6XS9vb12uz0UChWLRVInBvijOY2G6d7Vq6aRSqMRvr2F2Wxm7zbN3tro8vLS5/NpNJq+vj5BN7GXE6wTsgKBwNzcHJc3YtNoqOrm6F9vgy8nm81W/dc+MzPDMIzT6QyHw/l8vlQqxeNxrVbL3R5+Z2fHZrMVi0Wej6Kq21toNJquri673R6LxeLx+NPTE9sSI31ecpOkcCZDJo3+NBqmf4appv2hNBrewsKCz+fL5/O7u7ukz0VukhTOfjuTpog0GqZ/hqmmIRWn0YRib7OZSqUQTe00QZg6GTD+hTNqM2lKTKNhunffqmmqTaPVw10TfsOOzeDgIENNO43D85oQkwHjH1eSP5OG1JVGY9Xsn2GqaSpMo2HUG0KGYdirRHraaRw+Q4jPgAkaQpkzaT8OobLSaKya/TNMNU2FaTQR3t/fGYYxGAyUtNOEEpQB44+GTJqy0miofv8MU01TYRpNhPv7e4SQ1WqlpJ0mlLgMGB/EM2nKSqOh+v0zTDVNhWk0ES4uLhBCs7OzNLTTRBCXAfsRDZk0ZaXR+HTvqqtpKkyjCZXL5aLR6MDAwOLiIg3tNBHwGTDRhTMaMmmKSKNxqvtnfKppKkyj4TEM8/b2xhZwnp+fDw8Px8fHm5ubE4mEwWAg3k4TB58B4184Q/Rl0hSRRmPV7J/9WE1Dak2jVTs/Px8ZGWlvb9fpdE1NTej/TTNjY2PBYLBQKHDfSbadVo3nEkW9DBgjpHAmfyYNqSWNxtTvn+GraQyk0egn595R+TNpfIYQ0mgcNS9RAA6FmTRIo0kLhhCIAWk0CcEQ0ovyTBqk0aQCaTR60Z9JgzSaJOCVEADCYAgBIAyGEADCYAgBIEyCN2YymYzH42n8OErB7mlS8Y8cjUaPjo5In4UyZDIZditvIzRMY5/hj0QikLMDfxn7eZFGjtDoEAIAGgTXhAAQBkMIAGEwhAAQBkMIAGH/ARUanmrvCys/AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {},
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "###Loading the best autoencoder model"
      ],
      "metadata": {
        "id": "Dc1ncdIWe5dR"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "frauddetection_ae=load_model('frauddetectionmodel.h5')"
      ],
      "metadata": {
        "id": "W_wCogZxQWWh"
      },
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "frauddetection_ae.summary()"
      ],
      "metadata": {
        "id": "fjEWQvyiQWYx",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "399971b4-ee64-4229-9985-8a602f728c2d"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_3\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 73)]              0         \n",
            "                                                                 \n",
            " dense_4 (Dense)             (None, 37)                2738      \n",
            "                                                                 \n",
            " dense_5 (Dense)             (None, 18)                684       \n",
            "                                                                 \n",
            " dense_6 (Dense)             (None, 37)                703       \n",
            "                                                                 \n",
            " dense_7 (Dense)             (None, 73)                2774      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 6899 (26.95 KB)\n",
            "Trainable params: 6899 (26.95 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "for layer in frauddetection_ae.layers:\n",
        "  print(layer.name)"
      ],
      "metadata": {
        "id": "xPZcjIeXQWbg",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "f116da1c-edb9-443e-9714-05abbb7587ae"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "input_2\n",
            "dense_4\n",
            "dense_5\n",
            "dense_6\n",
            "dense_7\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "frauddetection_ae.layers[4].name"
      ],
      "metadata": {
        "id": "S5I95iH2ffnZ",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "482a39ce-b4eb-4663-f4a1-13d0d75b9083"
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'dense_7'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "frauddetection_ae.layers[4].output"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "YtEIRQF76yQF",
        "outputId": "cfe6c03e-cfc6-4556-e592-88ee36b10229"
      },
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<KerasTensor: shape=(None, 73) dtype=float32 (created by layer 'dense_7')>"
            ]
          },
          "metadata": {},
          "execution_count": 27
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "class_model=Model(inputs = frauddetection_ae.input, outputs=frauddetection_ae.get_layer(frauddetection_ae.layers[4].name).output)"
      ],
      "metadata": {
        "id": "GxZnbFNrfWgu"
      },
      "execution_count": 28,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "class_model.summary()"
      ],
      "metadata": {
        "id": "boH6h_MwfWi-",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "44941a35-6859-41dc-cb99-fc7a0347c759"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Model: \"model_4\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " input_2 (InputLayer)        [(None, 73)]              0         \n",
            "                                                                 \n",
            " dense_4 (Dense)             (None, 37)                2738      \n",
            "                                                                 \n",
            " dense_5 (Dense)             (None, 18)                684       \n",
            "                                                                 \n",
            " dense_6 (Dense)             (None, 37)                703       \n",
            "                                                                 \n",
            " dense_7 (Dense)             (None, 73)                2774      \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 6899 (26.95 KB)\n",
            "Trainable params: 6899 (26.95 KB)\n",
            "Non-trainable params: 0 (0.00 Byte)\n",
            "_________________________________________________________________\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "predictions= frauddetection_ae.predict(x_test)\n",
        "\n",
        "mse= np.mean(np.power(x_test - predictions,2),axis=1)\n",
        "\n",
        "loss = pd.DataFrame({'error':mse,\n",
        "                     'actuals':y_test})\n",
        "loss.describe()"
      ],
      "metadata": {
        "id": "diNv96EffWlb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 317
        },
        "outputId": "243cf2f9-bf14-4a4b-e3c9-56e761a5c016"
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "194/194 [==============================] - 0s 1ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "             error      actuals\n",
              "count  6177.000000  6177.000000\n",
              "mean      0.116729     0.491339\n",
              "std       0.193319     0.499965\n",
              "min       0.000573     0.000000\n",
              "25%       0.035480     0.000000\n",
              "50%       0.054717     0.000000\n",
              "75%       0.091335     1.000000\n",
              "max       0.991975     1.000000"
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-3a94994e-cdc3-43c5-805c-012f7144814d\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>error</th>\n",
              "      <th>actuals</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>count</th>\n",
              "      <td>6177.000000</td>\n",
              "      <td>6177.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>mean</th>\n",
              "      <td>0.116729</td>\n",
              "      <td>0.491339</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>std</th>\n",
              "      <td>0.193319</td>\n",
              "      <td>0.499965</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>min</th>\n",
              "      <td>0.000573</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>25%</th>\n",
              "      <td>0.035480</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>50%</th>\n",
              "      <td>0.054717</td>\n",
              "      <td>0.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>75%</th>\n",
              "      <td>0.091335</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>max</th>\n",
              "      <td>0.991975</td>\n",
              "      <td>1.000000</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-3a94994e-cdc3-43c5-805c-012f7144814d')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-3a94994e-cdc3-43c5-805c-012f7144814d button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-3a94994e-cdc3-43c5-805c-012f7144814d');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-68974851-755f-4879-a31d-d73ec18d7e37\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-68974851-755f-4879-a31d-d73ec18d7e37')\"\n",
              "            title=\"Suggest charts.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-68974851-755f-4879-a31d-d73ec18d7e37 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fpr,tpr,thresholds = roc_curve(loss.actuals,loss.error)\n",
        "\n",
        "roc_auc = auc(fpr,tpr)\n",
        "\n",
        "plt.title('Receiver Operating Characteristic')\n",
        "plt.plot(fpr, tpr, label ='AUC = %0.4f'%roc_auc)\n",
        "plt.legend(loc='lower right')\n",
        "plt.plot([0,1],[0,1],'r--')\n",
        "plt.ylabel('True Positive Rate')\n",
        "plt.xlabel('False Positive Rate')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "6Yva0QN5hL3h",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 480
        },
        "outputId": "79d2dd74-06ac-487f-dbcb-e1635d70013f"
      },
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "###Classification Model"
      ],
      "metadata": {
        "id": "I4rNiloHiV2y"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "frauddetection_cfn_model = Sequential()\n",
        "frauddetection_cfn_model.add(class_model)\n",
        "\n",
        "frauddetection_cfn_model.add(Dense(256,activation='tanh'))\n",
        "\n",
        "frauddetection_cfn_model.add(Dense(1,activation='sigmoid'))"
      ],
      "metadata": {
        "id": "LVQmA2U6gRkX"
      },
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "checkpoint2 = ModelCheckpoint('frauddetection_classification.h5', monitor = 'val_acc', verbose = 1, save_best_only=True, mode ='max')\n",
        "frauddetection_cfn_model.compile(loss=losses.mean_squared_error,\n",
        "                                 optimizer = 'adam',\n",
        "                                 metrics=['mse','acc'])\n",
        "\n",
        "fraudetection_history = frauddetection_cfn_model.fit(\n",
        "    x_train_supervised, y_train_supervised,\n",
        "    batch_size = 30,\n",
        "    shuffle = True,\n",
        "    callbacks = [checkpoint2],\n",
        "    epochs = 50,\n",
        "    validation_data=(x_test,y_test))"
      ],
      "metadata": {
        "id": "K_T0dstzitWm",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7ca8c443-1671-47a9-aa4b-16513706c7fa"
      },
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "411/412 [============================>.] - ETA: 0s - loss: 0.0239 - mse: 0.0239 - acc: 0.9715\n",
            "Epoch 1: val_acc improved from -inf to 0.99498, saving model to frauddetection_classification.h5\n",
            "412/412 [==============================] - 3s 5ms/step - loss: 0.0238 - mse: 0.0238 - acc: 0.9715 - val_loss: 0.0044 - val_mse: 0.0044 - val_acc: 0.9950\n",
            "Epoch 2/50\n",
            " 33/412 [=>............................] - ETA: 1s - loss: 0.0043 - mse: 0.0043 - acc: 0.9960"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "410/412 [============================>.] - ETA: 0s - loss: 0.0049 - mse: 0.0049 - acc: 0.9947\n",
            "Epoch 2: val_acc improved from 0.99498 to 0.99531, saving model to frauddetection_classification.h5\n",
            "412/412 [==============================] - 2s 6ms/step - loss: 0.0049 - mse: 0.0049 - acc: 0.9947 - val_loss: 0.0037 - val_mse: 0.0037 - val_acc: 0.9953\n",
            "Epoch 3/50\n",
            "409/412 [============================>.] - ETA: 0s - loss: 0.0036 - mse: 0.0036 - acc: 0.9957\n",
            "Epoch 3: val_acc did not improve from 0.99531\n",
            "412/412 [==============================] - 3s 7ms/step - loss: 0.0037 - mse: 0.0037 - acc: 0.9955 - val_loss: 0.0041 - val_mse: 0.0041 - val_acc: 0.9948\n",
            "Epoch 4/50\n",
            "406/412 [============================>.] - ETA: 0s - loss: 0.0026 - mse: 0.0026 - acc: 0.9968\n",
            "Epoch 4: val_acc improved from 0.99531 to 0.99563, saving model to frauddetection_classification.h5\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0026 - mse: 0.0026 - acc: 0.9968 - val_loss: 0.0035 - val_mse: 0.0035 - val_acc: 0.9956\n",
            "Epoch 5/50\n",
            "408/412 [============================>.] - ETA: 0s - loss: 0.0034 - mse: 0.0034 - acc: 0.9960\n",
            "Epoch 5: val_acc improved from 0.99563 to 0.99611, saving model to frauddetection_classification.h5\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 0.0034 - mse: 0.0034 - acc: 0.9960 - val_loss: 0.0038 - val_mse: 0.0038 - val_acc: 0.9961\n",
            "Epoch 6/50\n",
            "404/412 [============================>.] - ETA: 0s - loss: 0.0026 - mse: 0.0026 - acc: 0.9969\n",
            "Epoch 6: val_acc improved from 0.99611 to 0.99822, saving model to frauddetection_classification.h5\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 0.0027 - mse: 0.0027 - acc: 0.9968 - val_loss: 0.0014 - val_mse: 0.0014 - val_acc: 0.9982\n",
            "Epoch 7/50\n",
            "405/412 [============================>.] - ETA: 0s - loss: 0.0017 - mse: 0.0017 - acc: 0.9981\n",
            "Epoch 7: val_acc did not improve from 0.99822\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 0.0018 - mse: 0.0018 - acc: 0.9980 - val_loss: 0.0018 - val_mse: 0.0018 - val_acc: 0.9981\n",
            "Epoch 8/50\n",
            "399/412 [============================>.] - ETA: 0s - loss: 0.0020 - mse: 0.0020 - acc: 0.9974\n",
            "Epoch 8: val_acc improved from 0.99822 to 0.99854, saving model to frauddetection_classification.h5\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0019 - mse: 0.0019 - acc: 0.9975 - val_loss: 0.0015 - val_mse: 0.0015 - val_acc: 0.9985\n",
            "Epoch 9/50\n",
            "409/412 [============================>.] - ETA: 0s - loss: 0.0015 - mse: 0.0015 - acc: 0.9985\n",
            "Epoch 9: val_acc did not improve from 0.99854\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0015 - mse: 0.0015 - acc: 0.9985 - val_loss: 0.0022 - val_mse: 0.0022 - val_acc: 0.9974\n",
            "Epoch 10/50\n",
            "409/412 [============================>.] - ETA: 0s - loss: 0.0019 - mse: 0.0019 - acc: 0.9976\n",
            "Epoch 10: val_acc did not improve from 0.99854\n",
            "412/412 [==============================] - 3s 7ms/step - loss: 0.0019 - mse: 0.0019 - acc: 0.9976 - val_loss: 0.0015 - val_mse: 0.0015 - val_acc: 0.9982\n",
            "Epoch 11/50\n",
            "402/412 [============================>.] - ETA: 0s - loss: 0.0011 - mse: 0.0011 - acc: 0.9987\n",
            "Epoch 11: val_acc did not improve from 0.99854\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0011 - mse: 0.0011 - acc: 0.9987 - val_loss: 0.0019 - val_mse: 0.0019 - val_acc: 0.9981\n",
            "Epoch 12/50\n",
            "405/412 [============================>.] - ETA: 0s - loss: 0.0011 - mse: 0.0011 - acc: 0.9986\n",
            "Epoch 12: val_acc did not improve from 0.99854\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0012 - mse: 0.0012 - acc: 0.9985 - val_loss: 0.0240 - val_mse: 0.0240 - val_acc: 0.9704\n",
            "Epoch 13/50\n",
            "405/412 [============================>.] - ETA: 0s - loss: 0.0037 - mse: 0.0037 - acc: 0.9960\n",
            "Epoch 13: val_acc did not improve from 0.99854\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0036 - mse: 0.0036 - acc: 0.9961 - val_loss: 0.0027 - val_mse: 0.0027 - val_acc: 0.9971\n",
            "Epoch 14/50\n",
            "412/412 [==============================] - ETA: 0s - loss: 0.0023 - mse: 0.0023 - acc: 0.9976\n",
            "Epoch 14: val_acc did not improve from 0.99854\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 0.0023 - mse: 0.0023 - acc: 0.9976 - val_loss: 0.0076 - val_mse: 0.0076 - val_acc: 0.9919\n",
            "Epoch 15/50\n",
            "406/412 [============================>.] - ETA: 0s - loss: 0.0023 - mse: 0.0023 - acc: 0.9975\n",
            "Epoch 15: val_acc improved from 0.99854 to 0.99887, saving model to frauddetection_classification.h5\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0022 - mse: 0.0022 - acc: 0.9975 - val_loss: 0.0011 - val_mse: 0.0011 - val_acc: 0.9989\n",
            "Epoch 16/50\n",
            "409/412 [============================>.] - ETA: 0s - loss: 8.1561e-04 - mse: 8.1561e-04 - acc: 0.9992\n",
            "Epoch 16: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 3s 7ms/step - loss: 8.1039e-04 - mse: 8.1039e-04 - acc: 0.9992 - val_loss: 9.2259e-04 - val_mse: 9.2259e-04 - val_acc: 0.9989\n",
            "Epoch 17/50\n",
            "410/412 [============================>.] - ETA: 0s - loss: 5.9116e-04 - mse: 5.9116e-04 - acc: 0.9994\n",
            "Epoch 17: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 3s 7ms/step - loss: 6.2977e-04 - mse: 6.2977e-04 - acc: 0.9994 - val_loss: 0.0017 - val_mse: 0.0017 - val_acc: 0.9971\n",
            "Epoch 18/50\n",
            "402/412 [============================>.] - ETA: 0s - loss: 0.0017 - mse: 0.0017 - acc: 0.9981\n",
            "Epoch 18: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0017 - mse: 0.0017 - acc: 0.9981 - val_loss: 0.0020 - val_mse: 0.0020 - val_acc: 0.9977\n",
            "Epoch 19/50\n",
            "408/412 [============================>.] - ETA: 0s - loss: 0.0029 - mse: 0.0029 - acc: 0.9967\n",
            "Epoch 19: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 0.0029 - mse: 0.0029 - acc: 0.9968 - val_loss: 0.0047 - val_mse: 0.0047 - val_acc: 0.9950\n",
            "Epoch 20/50\n",
            "398/412 [===========================>..] - ETA: 0s - loss: 0.0025 - mse: 0.0025 - acc: 0.9971\n",
            "Epoch 20: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0025 - mse: 0.0025 - acc: 0.9971 - val_loss: 0.0034 - val_mse: 0.0034 - val_acc: 0.9963\n",
            "Epoch 21/50\n",
            "411/412 [============================>.] - ETA: 0s - loss: 0.0022 - mse: 0.0022 - acc: 0.9976\n",
            "Epoch 21: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0022 - mse: 0.0022 - acc: 0.9976 - val_loss: 0.0017 - val_mse: 0.0017 - val_acc: 0.9982\n",
            "Epoch 22/50\n",
            "409/412 [============================>.] - ETA: 0s - loss: 0.0014 - mse: 0.0014 - acc: 0.9985\n",
            "Epoch 22: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0014 - mse: 0.0014 - acc: 0.9985 - val_loss: 0.0027 - val_mse: 0.0027 - val_acc: 0.9969\n",
            "Epoch 23/50\n",
            "407/412 [============================>.] - ETA: 0s - loss: 0.0011 - mse: 0.0011 - acc: 0.9988\n",
            "Epoch 23: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 3s 7ms/step - loss: 0.0011 - mse: 0.0011 - acc: 0.9988 - val_loss: 0.0035 - val_mse: 0.0035 - val_acc: 0.9955\n",
            "Epoch 24/50\n",
            "395/412 [===========================>..] - ETA: 0s - loss: 0.0053 - mse: 0.0053 - acc: 0.9944\n",
            "Epoch 24: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 0.0051 - mse: 0.0051 - acc: 0.9946 - val_loss: 0.0020 - val_mse: 0.0020 - val_acc: 0.9979\n",
            "Epoch 25/50\n",
            "401/412 [============================>.] - ETA: 0s - loss: 0.0021 - mse: 0.0021 - acc: 0.9975\n",
            "Epoch 25: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0021 - mse: 0.0021 - acc: 0.9975 - val_loss: 0.0014 - val_mse: 0.0014 - val_acc: 0.9985\n",
            "Epoch 26/50\n",
            "403/412 [============================>.] - ETA: 0s - loss: 0.0012 - mse: 0.0012 - acc: 0.9988\n",
            "Epoch 26: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 0.0012 - mse: 0.0012 - acc: 0.9988 - val_loss: 9.3611e-04 - val_mse: 9.3611e-04 - val_acc: 0.9989\n",
            "Epoch 27/50\n",
            "411/412 [============================>.] - ETA: 0s - loss: 0.0014 - mse: 0.0014 - acc: 0.9985\n",
            "Epoch 27: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0014 - mse: 0.0014 - acc: 0.9985 - val_loss: 0.0021 - val_mse: 0.0021 - val_acc: 0.9977\n",
            "Epoch 28/50\n",
            "407/412 [============================>.] - ETA: 0s - loss: 0.0011 - mse: 0.0011 - acc: 0.9989\n",
            "Epoch 28: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0011 - mse: 0.0011 - acc: 0.9989 - val_loss: 0.0024 - val_mse: 0.0024 - val_acc: 0.9974\n",
            "Epoch 29/50\n",
            "412/412 [==============================] - ETA: 0s - loss: 0.0013 - mse: 0.0013 - acc: 0.9986\n",
            "Epoch 29: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 6ms/step - loss: 0.0013 - mse: 0.0013 - acc: 0.9986 - val_loss: 0.0013 - val_mse: 0.0013 - val_acc: 0.9987\n",
            "Epoch 30/50\n",
            "404/412 [============================>.] - ETA: 0s - loss: 0.0123 - mse: 0.0123 - acc: 0.9875\n",
            "Epoch 30: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 3s 7ms/step - loss: 0.0134 - mse: 0.0134 - acc: 0.9864 - val_loss: 0.0595 - val_mse: 0.0595 - val_acc: 0.9398\n",
            "Epoch 31/50\n",
            "396/412 [===========================>..] - ETA: 0s - loss: 0.0081 - mse: 0.0081 - acc: 0.9916\n",
            "Epoch 31: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0078 - mse: 0.0078 - acc: 0.9918 - val_loss: 0.0018 - val_mse: 0.0018 - val_acc: 0.9981\n",
            "Epoch 32/50\n",
            "409/412 [============================>.] - ETA: 0s - loss: 0.0014 - mse: 0.0014 - acc: 0.9985\n",
            "Epoch 32: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0014 - mse: 0.0014 - acc: 0.9985 - val_loss: 0.0025 - val_mse: 0.0025 - val_acc: 0.9972\n",
            "Epoch 33/50\n",
            "401/412 [============================>.] - ETA: 0s - loss: 0.0015 - mse: 0.0015 - acc: 0.9983\n",
            "Epoch 33: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0015 - mse: 0.0015 - acc: 0.9983 - val_loss: 0.0018 - val_mse: 0.0018 - val_acc: 0.9981\n",
            "Epoch 34/50\n",
            "402/412 [============================>.] - ETA: 0s - loss: 0.0034 - mse: 0.0034 - acc: 0.9961\n",
            "Epoch 34: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0033 - mse: 0.0033 - acc: 0.9962 - val_loss: 0.0020 - val_mse: 0.0020 - val_acc: 0.9979\n",
            "Epoch 35/50\n",
            "396/412 [===========================>..] - ETA: 0s - loss: 0.0012 - mse: 0.0012 - acc: 0.9987\n",
            "Epoch 35: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0012 - mse: 0.0012 - acc: 0.9988 - val_loss: 0.0011 - val_mse: 0.0011 - val_acc: 0.9987\n",
            "Epoch 36/50\n",
            "405/412 [============================>.] - ETA: 0s - loss: 0.0013 - mse: 0.0013 - acc: 0.9986\n",
            "Epoch 36: val_acc did not improve from 0.99887\n",
            "412/412 [==============================] - 3s 7ms/step - loss: 0.0013 - mse: 0.0013 - acc: 0.9986 - val_loss: 0.0021 - val_mse: 0.0021 - val_acc: 0.9977\n",
            "Epoch 37/50\n",
            "407/412 [============================>.] - ETA: 0s - loss: 0.0012 - mse: 0.0012 - acc: 0.9986\n",
            "Epoch 37: val_acc improved from 0.99887 to 0.99919, saving model to frauddetection_classification.h5\n",
            "412/412 [==============================] - 3s 7ms/step - loss: 0.0012 - mse: 0.0012 - acc: 0.9986 - val_loss: 8.5644e-04 - val_mse: 8.5644e-04 - val_acc: 0.9992\n",
            "Epoch 38/50\n",
            "397/412 [===========================>..] - ETA: 0s - loss: 5.9108e-04 - mse: 5.9108e-04 - acc: 0.9994\n",
            "Epoch 38: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 6.5100e-04 - mse: 6.5100e-04 - acc: 0.9994 - val_loss: 8.4922e-04 - val_mse: 8.4922e-04 - val_acc: 0.9992\n",
            "Epoch 39/50\n",
            "410/412 [============================>.] - ETA: 0s - loss: 6.4962e-04 - mse: 6.4962e-04 - acc: 0.9993\n",
            "Epoch 39: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 6.4704e-04 - mse: 6.4704e-04 - acc: 0.9994 - val_loss: 8.4071e-04 - val_mse: 8.4071e-04 - val_acc: 0.9992\n",
            "Epoch 40/50\n",
            "411/412 [============================>.] - ETA: 0s - loss: 0.0026 - mse: 0.0026 - acc: 0.9973\n",
            "Epoch 40: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0026 - mse: 0.0026 - acc: 0.9973 - val_loss: 0.0017 - val_mse: 0.0017 - val_acc: 0.9984\n",
            "Epoch 41/50\n",
            "402/412 [============================>.] - ETA: 0s - loss: 0.0025 - mse: 0.0025 - acc: 0.9970\n",
            "Epoch 41: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0025 - mse: 0.0025 - acc: 0.9970 - val_loss: 0.0045 - val_mse: 0.0045 - val_acc: 0.9937\n",
            "Epoch 42/50\n",
            "410/412 [============================>.] - ETA: 0s - loss: 0.0023 - mse: 0.0023 - acc: 0.9976\n",
            "Epoch 42: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 0.0023 - mse: 0.0023 - acc: 0.9976 - val_loss: 0.0014 - val_mse: 0.0014 - val_acc: 0.9982\n",
            "Epoch 43/50\n",
            "409/412 [============================>.] - ETA: 0s - loss: 5.8177e-04 - mse: 5.8177e-04 - acc: 0.9994\n",
            "Epoch 43: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 3s 8ms/step - loss: 5.7805e-04 - mse: 5.7805e-04 - acc: 0.9994 - val_loss: 0.0012 - val_mse: 0.0012 - val_acc: 0.9987\n",
            "Epoch 44/50\n",
            "405/412 [============================>.] - ETA: 0s - loss: 7.1468e-04 - mse: 7.1468e-04 - acc: 0.9993\n",
            "Epoch 44: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 7.0316e-04 - mse: 7.0316e-04 - acc: 0.9993 - val_loss: 7.1667e-04 - val_mse: 7.1667e-04 - val_acc: 0.9992\n",
            "Epoch 45/50\n",
            "404/412 [============================>.] - ETA: 0s - loss: 4.1594e-04 - mse: 4.1594e-04 - acc: 0.9996\n",
            "Epoch 45: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 4.0822e-04 - mse: 4.0822e-04 - acc: 0.9996 - val_loss: 7.3582e-04 - val_mse: 7.3582e-04 - val_acc: 0.9992\n",
            "Epoch 46/50\n",
            "401/412 [============================>.] - ETA: 0s - loss: 4.1569e-04 - mse: 4.1569e-04 - acc: 0.9996\n",
            "Epoch 46: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 4.0495e-04 - mse: 4.0495e-04 - acc: 0.9996 - val_loss: 7.3191e-04 - val_mse: 7.3191e-04 - val_acc: 0.9992\n",
            "Epoch 47/50\n",
            "398/412 [===========================>..] - ETA: 0s - loss: 3.3506e-04 - mse: 3.3506e-04 - acc: 0.9997\n",
            "Epoch 47: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 4.0492e-04 - mse: 4.0492e-04 - acc: 0.9996 - val_loss: 7.2965e-04 - val_mse: 7.2965e-04 - val_acc: 0.9992\n",
            "Epoch 48/50\n",
            "394/412 [===========================>..] - ETA: 0s - loss: 3.3845e-04 - mse: 3.3845e-04 - acc: 0.9997\n",
            "Epoch 48: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 5ms/step - loss: 4.0488e-04 - mse: 4.0488e-04 - acc: 0.9996 - val_loss: 7.2644e-04 - val_mse: 7.2644e-04 - val_acc: 0.9992\n",
            "Epoch 49/50\n",
            "409/412 [============================>.] - ETA: 0s - loss: 3.2578e-04 - mse: 3.2578e-04 - acc: 0.9997\n",
            "Epoch 49: val_acc did not improve from 0.99919\n",
            "412/412 [==============================] - 2s 4ms/step - loss: 4.0467e-04 - mse: 4.0467e-04 - acc: 0.9996 - val_loss: 7.1797e-04 - val_mse: 7.1797e-04 - val_acc: 0.9992\n",
            "Epoch 50/50\n",
            "405/412 [============================>.] - ETA: 0s - loss: 3.2940e-04 - mse: 3.2940e-04 - acc: 0.9997\n",
            "Epoch 50: val_acc improved from 0.99919 to 0.99935, saving model to frauddetection_classification.h5\n",
            "412/412 [==============================] - 3s 7ms/step - loss: 3.2409e-04 - mse: 3.2409e-04 - acc: 0.9997 - val_loss: 7.1386e-04 - val_mse: 7.1386e-04 - val_acc: 0.9994\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Predictions"
      ],
      "metadata": {
        "id": "5PUnzOn-kN1X"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "frauddetection_classification = load_model('frauddetection_classification.h5')\n",
        "\n",
        "y_test_preds = pd.DataFrame(frauddetection_classification.predict(x_test))\n",
        "y_test_pred = y_test_preds.apply(lambda x: x>=0.5)\n",
        "y_test_pred = y_test_pred.astype(int)"
      ],
      "metadata": {
        "id": "7hhToKAFitZI",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "7e7dbd45-ed88-40ed-a85e-313d073fd7ca"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "194/194 [==============================] - 0s 2ms/step\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "y=y_test.astype(int).reshape(6177,1)"
      ],
      "metadata": {
        "id": "Xyil72YKitbu"
      },
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "Mosaic Plot"
      ],
      "metadata": {
        "id": "uL09L587k48_"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "concantenatd = pd.DataFrame(np.concatenate((y_test_pred,y),axis=1))\n",
        "concantenatd.columns = ['Predicted','Actual']"
      ],
      "metadata": {
        "id": "eLgP26gfitd_"
      },
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "ax=mosaic(concantenatd,['Predicted','Actual'],gap=0.005)"
      ],
      "metadata": {
        "id": "P53faav_itgU",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        },
        "outputId": "52615daa-dd89-4842-9e4a-e828696c4442"
      },
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 3 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "###Confusion Matrix"
      ],
      "metadata": {
        "id": "_pXpiBqZlK1D"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "#confusion matrix\n",
        "cnf_matrix = confusion_matrix(y_test_pred,y_test)"
      ],
      "metadata": {
        "id": "6VaVxuBaiti5"
      },
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "cnf_matrix"
      ],
      "metadata": {
        "id": "56QNeOZ1itla",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "98d4658a-0cf2-444d-abcb-1b56961f9765"
      },
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[3138,    0],\n",
              "       [   4, 3035]])"
            ]
          },
          "metadata": {},
          "execution_count": 40
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from itertools import product"
      ],
      "metadata": {
        "id": "qJUnO8RZKzQB"
      },
      "execution_count": 43,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "classes=['Normal','Fraud']\n",
        "plt.imshow(cnf_matrix,interpolation='nearest',cmap=plt.cm.Blues)\n",
        "plt.title(\"Confusion Matrix\")\n",
        "plt.colorbar()\n",
        "tick_marks = np.arange(len(classes))\n",
        "plt.xticks(tick_marks,classes,rotation=45)\n",
        "plt.yticks(tick_marks,classes)\n",
        "\n",
        "thresh = cnf_matrix.max() / 2.\n",
        "for i,j in product(range(cnf_matrix.shape[0]), range(cnf_matrix.shape[1])):\n",
        "      plt.text(j, i, format(cnf_matrix[i,j], 'd'),\n",
        "               horizontalalignment = \"center\",\n",
        "               color = \"white\" if cnf_matrix[i,j]>thresh else \"black\")\n",
        "\n",
        "plt.tight_layout()\n",
        "plt.ylabel('True Label')\n",
        "plt.xlabel('Predicted Label')\n"
      ],
      "metadata": {
        "id": "OGrr0-FtgRn8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 521
        },
        "outputId": "c7424de3-d8b4-4bf4-a8f6-c844a261c4c2"
      },
      "execution_count": 50,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, -15.71263972600328, 'Predicted Label')"
            ]
          },
          "metadata": {},
          "execution_count": 50
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "pipe3 = Pipeline(steps=[\n",
        "    ('step1', preprocessor)\n",
        "])"
      ],
      "metadata": {
        "id": "m6MTyfAnmrpI"
      },
      "execution_count": 52,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# import pickle\n",
        "# pickle.dump(preprocessor,open('preprocessor.pkl','wb'))"
      ],
      "metadata": {
        "id": "74LRDKyRfWnt"
      },
      "execution_count": 55,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "frauddetection_classification.save(\"AutoEncoder.h5\")"
      ],
      "metadata": {
        "id": "xj8-kSmqmrrt",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0ba207fd-1293-47e1-ba9d-ed887be67a9a"
      },
      "execution_count": 51,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3079: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n",
            "  saving_api.save_model(\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fraudulent_samples = train_new[train['is_fraud'] == 1]\n",
        "fraudulent_samples=fraudulent_samples[0:5]\n",
        "fraudulent_samples"
      ],
      "metadata": {
        "id": "DtZey33Omrt8",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 226
        },
        "outputId": "1d60a1a1-1ab8-4cfe-a2fe-3f7788cb9417"
      },
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "           category state     amt    zip      lat     long  city_pop  \\\n",
              "2449    grocery_pos    NC  281.06  28611  35.9946 -81.7266       885   \n",
              "2472  gas_transport    TX   11.52  78208  29.4400 -98.4590   1595797   \n",
              "2523    grocery_pos    TX  276.31  78208  29.4400 -98.4590   1595797   \n",
              "2546  gas_transport    NC    7.03  28611  35.9946 -81.7266       885   \n",
              "2553    grocery_pos    TX  275.73  78208  29.4400 -98.4590   1595797   \n",
              "\n",
              "      merch_lat  merch_long  age  hour  day  month  is_fraud  \n",
              "2449  36.430124  -81.179483   35     1    2      1       1.0  \n",
              "2472  29.819364  -99.142791   63     1    2      1       1.0  \n",
              "2523  29.273085  -98.836360   63     3    2      1       1.0  \n",
              "2546  35.909292  -82.091010   35     3    2      1       1.0  \n",
              "2553  29.786426  -98.683410   63     3    2      1       1.0  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-ec37fe6f-adb1-4726-a3cd-a473dbfe45a5\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>category</th>\n",
              "      <th>state</th>\n",
              "      <th>amt</th>\n",
              "      <th>zip</th>\n",
              "      <th>lat</th>\n",
              "      <th>long</th>\n",
              "      <th>city_pop</th>\n",
              "      <th>merch_lat</th>\n",
              "      <th>merch_long</th>\n",
              "      <th>age</th>\n",
              "      <th>hour</th>\n",
              "      <th>day</th>\n",
              "      <th>month</th>\n",
              "      <th>is_fraud</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2449</th>\n",
              "      <td>grocery_pos</td>\n",
              "      <td>NC</td>\n",
              "      <td>281.06</td>\n",
              "      <td>28611</td>\n",
              "      <td>35.9946</td>\n",
              "      <td>-81.7266</td>\n",
              "      <td>885</td>\n",
              "      <td>36.430124</td>\n",
              "      <td>-81.179483</td>\n",
              "      <td>35</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2472</th>\n",
              "      <td>gas_transport</td>\n",
              "      <td>TX</td>\n",
              "      <td>11.52</td>\n",
              "      <td>78208</td>\n",
              "      <td>29.4400</td>\n",
              "      <td>-98.4590</td>\n",
              "      <td>1595797</td>\n",
              "      <td>29.819364</td>\n",
              "      <td>-99.142791</td>\n",
              "      <td>63</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2523</th>\n",
              "      <td>grocery_pos</td>\n",
              "      <td>TX</td>\n",
              "      <td>276.31</td>\n",
              "      <td>78208</td>\n",
              "      <td>29.4400</td>\n",
              "      <td>-98.4590</td>\n",
              "      <td>1595797</td>\n",
              "      <td>29.273085</td>\n",
              "      <td>-98.836360</td>\n",
              "      <td>63</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2546</th>\n",
              "      <td>gas_transport</td>\n",
              "      <td>NC</td>\n",
              "      <td>7.03</td>\n",
              "      <td>28611</td>\n",
              "      <td>35.9946</td>\n",
              "      <td>-81.7266</td>\n",
              "      <td>885</td>\n",
              "      <td>35.909292</td>\n",
              "      <td>-82.091010</td>\n",
              "      <td>35</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2553</th>\n",
              "      <td>grocery_pos</td>\n",
              "      <td>TX</td>\n",
              "      <td>275.73</td>\n",
              "      <td>78208</td>\n",
              "      <td>29.4400</td>\n",
              "      <td>-98.4590</td>\n",
              "      <td>1595797</td>\n",
              "      <td>29.786426</td>\n",
              "      <td>-98.683410</td>\n",
              "      <td>63</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "      <td>1.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-ec37fe6f-adb1-4726-a3cd-a473dbfe45a5')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-ec37fe6f-adb1-4726-a3cd-a473dbfe45a5 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-ec37fe6f-adb1-4726-a3cd-a473dbfe45a5');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-0b6178ec-98b4-4732-a9d9-a05ec7fac2a8\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-0b6178ec-98b4-4732-a9d9-a05ec7fac2a8')\"\n",
              "            title=\"Suggest charts.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-0b6178ec-98b4-4732-a9d9-a05ec7fac2a8 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "fraudulent_samples=fraudulent_samples.iloc[:, :-1]\n",
        "fraudulent_samples"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 226
        },
        "id": "WVzACNvDyLdF",
        "outputId": "742a99f0-2dac-4719-87dd-231713d1a27b"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "           category state     amt    zip      lat     long  city_pop  \\\n",
              "2449    grocery_pos    NC  281.06  28611  35.9946 -81.7266       885   \n",
              "2472  gas_transport    TX   11.52  78208  29.4400 -98.4590   1595797   \n",
              "2523    grocery_pos    TX  276.31  78208  29.4400 -98.4590   1595797   \n",
              "2546  gas_transport    NC    7.03  28611  35.9946 -81.7266       885   \n",
              "2553    grocery_pos    TX  275.73  78208  29.4400 -98.4590   1595797   \n",
              "\n",
              "      merch_lat  merch_long  age  hour  day  month  \n",
              "2449  36.430124  -81.179483   35     1    2      1  \n",
              "2472  29.819364  -99.142791   63     1    2      1  \n",
              "2523  29.273085  -98.836360   63     3    2      1  \n",
              "2546  35.909292  -82.091010   35     3    2      1  \n",
              "2553  29.786426  -98.683410   63     3    2      1  "
            ],
            "text/html": [
              "\n",
              "  <div id=\"df-a7e7ec47-7e35-4e09-b647-dfcb9ed138c9\" class=\"colab-df-container\">\n",
              "    <div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>category</th>\n",
              "      <th>state</th>\n",
              "      <th>amt</th>\n",
              "      <th>zip</th>\n",
              "      <th>lat</th>\n",
              "      <th>long</th>\n",
              "      <th>city_pop</th>\n",
              "      <th>merch_lat</th>\n",
              "      <th>merch_long</th>\n",
              "      <th>age</th>\n",
              "      <th>hour</th>\n",
              "      <th>day</th>\n",
              "      <th>month</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>2449</th>\n",
              "      <td>grocery_pos</td>\n",
              "      <td>NC</td>\n",
              "      <td>281.06</td>\n",
              "      <td>28611</td>\n",
              "      <td>35.9946</td>\n",
              "      <td>-81.7266</td>\n",
              "      <td>885</td>\n",
              "      <td>36.430124</td>\n",
              "      <td>-81.179483</td>\n",
              "      <td>35</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2472</th>\n",
              "      <td>gas_transport</td>\n",
              "      <td>TX</td>\n",
              "      <td>11.52</td>\n",
              "      <td>78208</td>\n",
              "      <td>29.4400</td>\n",
              "      <td>-98.4590</td>\n",
              "      <td>1595797</td>\n",
              "      <td>29.819364</td>\n",
              "      <td>-99.142791</td>\n",
              "      <td>63</td>\n",
              "      <td>1</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2523</th>\n",
              "      <td>grocery_pos</td>\n",
              "      <td>TX</td>\n",
              "      <td>276.31</td>\n",
              "      <td>78208</td>\n",
              "      <td>29.4400</td>\n",
              "      <td>-98.4590</td>\n",
              "      <td>1595797</td>\n",
              "      <td>29.273085</td>\n",
              "      <td>-98.836360</td>\n",
              "      <td>63</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2546</th>\n",
              "      <td>gas_transport</td>\n",
              "      <td>NC</td>\n",
              "      <td>7.03</td>\n",
              "      <td>28611</td>\n",
              "      <td>35.9946</td>\n",
              "      <td>-81.7266</td>\n",
              "      <td>885</td>\n",
              "      <td>35.909292</td>\n",
              "      <td>-82.091010</td>\n",
              "      <td>35</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2553</th>\n",
              "      <td>grocery_pos</td>\n",
              "      <td>TX</td>\n",
              "      <td>275.73</td>\n",
              "      <td>78208</td>\n",
              "      <td>29.4400</td>\n",
              "      <td>-98.4590</td>\n",
              "      <td>1595797</td>\n",
              "      <td>29.786426</td>\n",
              "      <td>-98.683410</td>\n",
              "      <td>63</td>\n",
              "      <td>3</td>\n",
              "      <td>2</td>\n",
              "      <td>1</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-a7e7ec47-7e35-4e09-b647-dfcb9ed138c9')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-a7e7ec47-7e35-4e09-b647-dfcb9ed138c9 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-a7e7ec47-7e35-4e09-b647-dfcb9ed138c9');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-aca6ca4d-8639-4c6c-b029-372a1226c8a0\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-aca6ca4d-8639-4c6c-b029-372a1226c8a0')\"\n",
              "            title=\"Suggest charts.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-aca6ca4d-8639-4c6c-b029-372a1226c8a0 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "    </div>\n",
              "  </div>\n"
            ]
          },
          "metadata": {},
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "data=preprocessor.transform(fraudulent_samples)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qT0YNQssy3Jx",
        "outputId": "4b98ca17-8593-46a8-d0d2-a7312a797c8c"
      },
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/base.py:432: UserWarning: X has feature names, but StandardScaler was fitted without feature names\n",
            "  warnings.warn(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/base.py:432: UserWarning: X has feature names, but OneHotEncoder was fitted without feature names\n",
            "  warnings.warn(\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "frauddetection_classification.predict(data)"
      ],
      "metadata": {
        "id": "J-_xWwAWmrwm",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "621d06b8-7091-475a-c53e-fc2e128fbe38"
      },
      "execution_count": 68,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 35ms/step\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[0.9999737],\n",
              "       [1.       ],\n",
              "       [1.       ],\n",
              "       [0.9999951],\n",
              "       [1.       ]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 68
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [],
      "metadata": {
        "id": "5rbtclQexw9L"
      }
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "_qrtcjlTmrzG"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "eUVsJMPJmr1Y"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import pickle\n",
        "with open('logisticRegression.pkl', 'rb') as model_file:\n",
        "    logistic_model = pickle.load(model_file)\n",
        "with open('RandomForest.pkl', 'rb') as model_file:\n",
        "    random_forest = pickle.load(model_file)\n",
        "with open('SVC.pkl', 'rb') as model_file:\n",
        "    svc = pickle.load(model_file)"
      ],
      "metadata": {
        "id": "Lm0mKvmJmr33"
      },
      "execution_count": 33,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "with open('pipe2.pkl', 'rb') as model_file:\n",
        "    random_forest = pickle.load(model_file)"
      ],
      "metadata": {
        "id": "nuY9XCkGnZzo"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "AutoEncoder_model =load_model('AutoEncoder.h5')"
      ],
      "metadata": {
        "id": "IIGIy-7Ocpb8"
      },
      "execution_count": 35,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.ensemble import VotingClassifier\n",
        "from sklearn.metrics import accuracy_score"
      ],
      "metadata": {
        "id": "iu35ZFH5c910"
      },
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# Create a Voting Classifier\n",
        "ensemble_model = VotingClassifier(estimators=[\n",
        "    ('linear', logistic_model),\n",
        "    ('random_forest', random_forest),\n",
        "    ('svc', svc)\n",
        "], voting='hard')\n",
        "\n",
        "# Fit the ensemble model on the training data\n",
        "ensemble_model.fit(creditCardFeatures_sampled, creditCardLabels_sampled)\n",
        "\n",
        "# Make predictions on the validation set\n",
        "ensemble_predictions = ensemble_model.predict(x_test)\n",
        "\n",
        "# Evaluate the ensemble model\n",
        "accuracy = accuracy_score(y_test, ensemble_predictions)\n",
        "print(\"Ensemble Model Accuracy:\", accuracy)"
      ],
      "metadata": {
        "id": "yt5oLu81QWgQ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "78dcd10a-5f6b-41c6-e882-419fa55f1a27"
      },
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Ensemble Model Accuracy: 0.9698882952889752\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "ensemble = Pipeline(steps=[\n",
        "    ('step1', preprocessor),\n",
        "    ('step2', ensemble_model)\n",
        "])"
      ],
      "metadata": {
        "id": "OZxbSZBfpe0q"
      },
      "execution_count": 40,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import pickle\n",
        "pickle.dump(ensemble,open('ensemble.pkl','wb'))"
      ],
      "metadata": {
        "id": "rR-Oo2Pqpo3p"
      },
      "execution_count": 43,
      "outputs": []
    }
  ]
}